So I always thought there was no diminishing return on resistance, and I remember I tested that in the past.
But after talking about it with a friend and looking at resistance numbers, now it seems there is a (fairly small) diminishing return afterall.
I did some tests with these results.
Build
109 from defense
10 from tank role
19 per defiance stack
Results per defiance stack
Expected 119
You lose 690 (1500) Fire Damage from Blue Beam.
Calculated 1500/690-1=1.1739
Expected 119+19=138
You lose 636 (1500) Fire Damage from Blue Beam.
Calculated 1500/636-1=1.3584
Expected 119+19*2=157
You lose 590 (1500) Fire Damage from Blue Beam.
Calculated 1500/5902-1=1.5423
Expected 119+19*3=176
You lose 551 (1500) Fire Damage from Blue Beam.
Calculated 1500/5512-1=1.7223
Expected 119+19*4=195
You lose 517 (1500) Fire Damage from Blue Beam.
Calculated 1500/5172-1=1.90135
Expected 119+19*5=214
You lose 488 (1500) Fire Damage from Blue Beam.
Calculated 1500/4882-1=2.0737
Expected 119+19*6=233
You lose 462 (1500) Fire Damage from Blue Beam.
Calculated 1500/462-1=2.24675
Just to be sure it's not defiance I also tested with cold resist mods.
Build
100 from defense
50 per cold resist mod
Results per mod
Expected 1
X deals 5378 (10684) Cold Damage to you with Icicle Spear.
Calculated 10684/5378-1=0.986
Expected 1.5
X deals 2149 (5301) Cold Damage to you with Icicle Spear.
Calculated 5301/2149-1=1.466
Expected 2
X deals 1719 (5039) Cold Damage to you with Icicle Spear.
Calculated 5039/1719-1=1.931
Is this something new, or has it always worked like this?
Comments
Defense was 443.1, which gave a tooltip value of 104%, and tank role. Damage from blue laser was 705, which adds up to 112.7%. In theory that could be something like 704.1 damage (rounded up) with damage resistance 103.03% (rounded up), but that seems not likely.
Might turn on combat logging, as I recall combatlog.log actually shows damage to tenths.
Epic Stronghold
Block timing explained
So predicted values based on tooltips will mostly skew towards being higher than actuals.
PARSER USAGE GUIDE: click here for an easy how-to ٩(๑・ิᴗ・ิ)۶٩(・ิᴗ・ิ๑)۶
Spectre beats Eido in an SG run: VIDEO HERE
- Be safe and have fun, champs - for science!
But even assuming the numbers are actually 99.1+49.1*2 that would never come close to an effective 193.1% resistance.
Epic Stronghold
Block timing explained
PARSER USAGE GUIDE: click here for an easy how-to ٩(๑・ิᴗ・ิ)۶٩(・ิᴗ・ิ๑)۶
Spectre beats Eido in an SG run: VIDEO HERE
The 342% resistance test is also quite interesting since at that point it really starts to become significant, and 342% is not unreasonable at all when you consider AoRP.
If you take a 10000 base damage hit with 342% resistance, you'd expect 2262 damage but you actually take 2399.
I 've used some polynomial regression website to find a function that fits the current data points fairly well.
Effective Resistance = -0.03*Resistance^2 + 1.035*Resistance - 0.01
And updated a spreadsheet I made a while ago to easily calculate "effective health".
https://www.dropbox.com/s/tt0gf0rwpv7giyq/mitigation calc.ods?dl=0
Epic Stronghold
Block timing explained
But something else I've found which is interesting, block only works for 83.33% of the normal block resistance for AoE.
Epic Stronghold
Block timing explained
Epic Stronghold
Block timing explained
It seems elusive does block resist.
So to recalculate the first AoE hit on each block At least it shows block resist does not have any DR.