There are no diminishing returns on armor. That's been a common misunderstanding of basic math that has existed since vanilla.Darkmoon card is legit on a few fights but i dont think its the end all be all trinket that the warrior discord splooges over every time trinkets are brought up (same with autoblocker). once you are in the 7k Armor range DR makes the armor less effective than if you are lower, you can compare it for yourself by looking at the damage reduction you get from heavy hide buff at your current armor value vs what you would get if you were higher when it proc'd.
Case A: Base of 1000 armor
Dmg reduction = 11.92%
With Heavy Hide proc: 1000 + 3000 = 4000 armor
Dmg reduction = 35.12%
Increase of 23.2%
Case B: Base of 7000 armor
Dmg reduction = 48.64%
With Heavy Hide proc: 7000 + 3000 = 10,000 armor
Dmg reduction = 57.5%
Increase of 8.86%
8.86% is less than 23.2%, so that's diminishing returns, right? NO, BAD MATH! To know how the value of armor really scales you need to compare the actual AMOUNT of dmg reduced in both cases. Let's take an unmitigated hit of 1,000,000 as an example:
Case A: 1000 armor = 11.92% reduction
880,800 damage taken
With proc: 4000 armor = 35.12%
648,800 damage taken
232,000 damage reduced, or 26.34% of non-proc hit.
Case B: 7000 armor = 48.64%
513,600 damage taken
With proc: 10,000 armor = 57.5%
425,000 damage taken
88,600 damage reduced, or 17.25% of non-proc hit.
Oh wait, 17.25% is less than 26.34%, shit. I guess there are diminishing returns after all, right? NO, STILL BAD MATH! HOLD THE FUCK ON WE'RE ALMOST THERE.
We're one step away from working with numbers that are actually directly indicative of a tank's effectiveness to stay alive. Let's combine an example max HP of 5,000,000 with the previous results and see what happens.
Case A: 1000 armor = 880,800 damage taken per hit
5.68 hits until death, (or 11.35 seconds at 2 sec swing timer, or 5,676,658 effective health, whichever you want to use)
proc: 4000 armor = 648,800 dmg per hit
7.71 hits to death, (15.41 sec, 7,706,535 EH)
2.03 hit increase, (4.06 sec, 2,029,877 EH)
Case B: 7000 armor = 513,600 dmg per hit
9.74 hits to death, (19.47 sec, 9,735,202 EH)
proc: 10,000 armor = 425,000 dmg per hit
11.76 hits to death, (23.53 sec, 11,764,706 EH)
2.02 hit increase, (4.06 sec, 2,029,504 EH)
The minor discrepancy between these two cases is entirely due to rounding during calculations (fuck sig figs). The actual results are EXACTLY the same.
Dmg reduction = 11.92%
With Heavy Hide proc: 1000 + 3000 = 4000 armor
Dmg reduction = 35.12%
Increase of 23.2%
Case B: Base of 7000 armor
Dmg reduction = 48.64%
With Heavy Hide proc: 7000 + 3000 = 10,000 armor
Dmg reduction = 57.5%
Increase of 8.86%
8.86% is less than 23.2%, so that's diminishing returns, right? NO, BAD MATH! To know how the value of armor really scales you need to compare the actual AMOUNT of dmg reduced in both cases. Let's take an unmitigated hit of 1,000,000 as an example:
Case A: 1000 armor = 11.92% reduction
880,800 damage taken
With proc: 4000 armor = 35.12%
648,800 damage taken
232,000 damage reduced, or 26.34% of non-proc hit.
Case B: 7000 armor = 48.64%
513,600 damage taken
With proc: 10,000 armor = 57.5%
425,000 damage taken
88,600 damage reduced, or 17.25% of non-proc hit.
Oh wait, 17.25% is less than 26.34%, shit. I guess there are diminishing returns after all, right? NO, STILL BAD MATH! HOLD THE FUCK ON WE'RE ALMOST THERE.
We're one step away from working with numbers that are actually directly indicative of a tank's effectiveness to stay alive. Let's combine an example max HP of 5,000,000 with the previous results and see what happens.
Case A: 1000 armor = 880,800 damage taken per hit
5.68 hits until death, (or 11.35 seconds at 2 sec swing timer, or 5,676,658 effective health, whichever you want to use)
proc: 4000 armor = 648,800 dmg per hit
7.71 hits to death, (15.41 sec, 7,706,535 EH)
2.03 hit increase, (4.06 sec, 2,029,877 EH)
Case B: 7000 armor = 513,600 dmg per hit
9.74 hits to death, (19.47 sec, 9,735,202 EH)
proc: 10,000 armor = 425,000 dmg per hit
11.76 hits to death, (23.53 sec, 11,764,706 EH)
2.02 hit increase, (4.06 sec, 2,029,504 EH)
The minor discrepancy between these two cases is entirely due to rounding during calculations (fuck sig figs). The actual results are EXACTLY the same.
Last edited:
- 1