test content
Logo
What is the Arc Client?
Install Arc
Disruptor Percent Uptime
jbmaverick
Member Posts: 935 Arc User
I'm trying to figure out the percent uptime of disruptor procs on standard weapons so I can compare Fleet Plasma weapons (ACCx2, DMGx2) to Romulan Reputation weapons (ACCx2), but I'm afraid my statistics skills aren't nearly good enough to go about it. This is strictly for PvE, so no worries about 80% of PvP targets having high plasma resists, and I know it's not really a big deal to min-max PvE to this degree. I'm just curious if the fleet DMGx2 modifier and slightly higher rarity is better than the disruptor effect of Romulan weapons. Here's what I've got so far:
Most (if not all) disruptor weapons and hybrids have a 2.5% chance per weapon activation cycle to apply their debuff. Obviously the total chance of applying the debuff will depend on the number and type of weapons fired.
The chance of the debuff occurring per cycle (assuming synchronized weapon cycles) is (Chance of weapon 1 proccing) + (Chance of weapon 1 not proccing * Chance of weapon 2 proccing) + (Chance of weapon 1 not proccing * Chance of weapon 2 not proccing * Chance of weapon 3 proccing) etc. until you reach the number of weapons with the disruptor proc that you are firing. The reasoning for the complicated math is because disruptor debuffs do not stack, so if Weapon 1 has applied the effect, it doesn't matter if Weapons 2 through X apply it, etc. beacuse it doesn't change the overall effect . Similarly, we don't care if Weapon 2 procs if Weapon 1 already did, so we only calculate the chance for Weapon 2 based on the assumption that Weapon 1 failed, etc. Feel free to correct me on this, it's been awhile since I've done much statistical analysis and I may be forgetting something.
So, for example, 5 weapons would have (ignoring accuracy and defense calculations):
(0.025) + (0.975 * 0.025) + (0.975^2 * 0.025) + (0.975^3 * 0.025) + (0.974^4 * 0.025) = about 11.9% chance of applying a disruptor proc to your target, assuming all 5 weapons can hit (cannons + turrets, broadside arrays, etc.)
This is where I'm not sure how to progress. The debuff lasts 15 seconds, meaning you can get 3 beam or 5 cannon firing cycles in before the debuff expires, and in that 15 seconds the debuff could be reapplied, starting a new 15 second timer until expiration. How would one go about figuring out the average percent uptime of the disruptor debuff?
I suppose a simpler solution would be to use a DPS program to compare the 2 weapon sets in a controlled environment, but I'm still interested in the numerical results of this particular question.
I am aware that disruptors are being debuffed/fixed, I only did damage calculations based on a fixed -10.2 damage resistance and assumed positive resistances at several stages since I don't think anyone really knows what NPC resistances are set at specifically.
Most (if not all) disruptor weapons and hybrids have a 2.5% chance per weapon activation cycle to apply their debuff. Obviously the total chance of applying the debuff will depend on the number and type of weapons fired.
The chance of the debuff occurring per cycle (assuming synchronized weapon cycles) is (Chance of weapon 1 proccing) + (Chance of weapon 1 not proccing * Chance of weapon 2 proccing) + (Chance of weapon 1 not proccing * Chance of weapon 2 not proccing * Chance of weapon 3 proccing) etc. until you reach the number of weapons with the disruptor proc that you are firing. The reasoning for the complicated math is because disruptor debuffs do not stack, so if Weapon 1 has applied the effect, it doesn't matter if Weapons 2 through X apply it, etc. beacuse it doesn't change the overall effect . Similarly, we don't care if Weapon 2 procs if Weapon 1 already did, so we only calculate the chance for Weapon 2 based on the assumption that Weapon 1 failed, etc. Feel free to correct me on this, it's been awhile since I've done much statistical analysis and I may be forgetting something.
So, for example, 5 weapons would have (ignoring accuracy and defense calculations):
(0.025) + (0.975 * 0.025) + (0.975^2 * 0.025) + (0.975^3 * 0.025) + (0.974^4 * 0.025) = about 11.9% chance of applying a disruptor proc to your target, assuming all 5 weapons can hit (cannons + turrets, broadside arrays, etc.)
This is where I'm not sure how to progress. The debuff lasts 15 seconds, meaning you can get 3 beam or 5 cannon firing cycles in before the debuff expires, and in that 15 seconds the debuff could be reapplied, starting a new 15 second timer until expiration. How would one go about figuring out the average percent uptime of the disruptor debuff?
I suppose a simpler solution would be to use a DPS program to compare the 2 weapon sets in a controlled environment, but I'm still interested in the numerical results of this particular question.
I am aware that disruptors are being debuffed/fixed, I only did damage calculations based on a fixed -10.2 damage resistance and assumed positive resistances at several stages since I don't think anyone really knows what NPC resistances are set at specifically.
The universe has a wonderful sense of humor. The trick is learning how to take a joke.
Post edited by jbmaverick on
0
Comments
4 Beam Weapons
About 26.2%
5 Beam Weapons
About 31.6%
6 Beam Weapons
About 36.6%
7 Beam Weapons
About 41.2%
8 Beam Weapons
About 45.5%
4 Cannon Weapons
About 39.7%
5 Cannon Weapons
About 46.9%
6 Cannon Weapons
About 53.2%
7 Cannon Weapons
About 58.8%
8 Cannon Weapons
About 63.7%
This all assumes, of course, that you can keep all of your weapons on target without delaying any firing cycles, which isn't always going to happen. Also, this only applies to Disruptor procs, as it's the only non-stacking, non-cooldown debuff weapon proc with a duration. Phasers are affected by Phaser System Offline Immunity, Plasma can stack so it does matter if multiple weapons proc it, Tetryon and Polaron are just added damage (shield and power respectively) so it also matters if multiple weapons proc it.
Again, feel free to correct me if I've got this wrong, it still feels like I might be missing something.
The universe has a wonderful sense of humor. The trick is learning how to take a joke.
Disruptors,
Nanite Disruptors,
Hybrid Plasma Disruptors,
and Romulan Plasma with a Disruptor proc...
Then there were the two questions of whether a second proc overwrites/extends a previous proc as well as whether a stronger proc overwrites a weaker proc. That's before getting into the role accuracy/defense and To-Hit would play.
I've got a guy that's been running Rom Plasma since he was able to get them, but I've never really paid attention to the proc uptime on the Disruptor proc. I had another guy that's run regular Disruptors forever, but again - never really paid attention. With the recent announcement that the Disruptor proc was being fixed, I actually decided to make the move on two of my guys from various Polarons to various Disruptors.
So I'm curious as to where this overall discussion will lead...
It looks like the numbers are based off of standard probability (which is why I avoided this conversation because of my earlier questions)...
4 Beams 2.5% chance each, (.975^4)...15s period, 3 cycles...((.975^4)^3).
But do they overwrite/extend, etc, etc, etc - which could throw everything off, no?
The universe has a wonderful sense of humor. The trick is learning how to take a joke.
If the proc occurs on shot 3 of the 2nd cycle, then there will be a gap if it happens later than that in the next set of cycles.
Besides, the uptimes I've found assume you're never missing and your target is always in the firing arc of all your weapons capable of causing the effect. In actual combat, PvE or otherwise, this will rarely be the case, so this is more of a max uptime calculation than an effective one.
The universe has a wonderful sense of humor. The trick is learning how to take a joke.
There is a simpler formula for this. The probability that at least one of 5 weapons procs during one firing cycle is 1 - (1 - 0.025)^5.
Does the above assume that the weapons fire in sync? The calculation is complicated by the fact that weapons do not fire in sync, but in a staggered fashion. For example, if you fire 8 beam arrays, it's possible that the 1st beam array fires 3 times within a 15-second window, but the 8th one fires only twice. Does anyone know what the delay between firing consecutive weapons is?
And again, the final result is a maximum uptime percentage given perfect conditions, which is useful to see how much a Disruptor proc could affect your average damage at max. There are a lot of complications that could reduce the uptime, but nothing is going to increase it beyond what I found, assuming all the math is correct.
The universe has a wonderful sense of humor. The trick is learning how to take a joke.
If you just want an upper bound, then yes, your numbers are correct. But there is a simpler way to calculate them. My comments are in red.
For example, I was trying to test something else when I noticed the following while doing Empire Defense - Federation (each was vs. 2 Battleships, 2 Cruisers, and 4 Frigates (Light Cruisers)):
2x Nanite Disruptor Beam Arrays
0 Disruptor Procs
3 Nanite Cloud Procs
2x Hybrid Plasma-Disruptor Beam Arrays
7 Disruptor Procs
7 Plasma Procs
(they were all dual procs - both Disruptor and Plasma at the same time for each of the 7 times)
1x Hybrid Plasma-Disruptor Beam Array & 1x Polarized Disruptor Beam Array
6 Disruptor Procs
2 Plasma Procs
4 Polaron Procs
(again, dual procs - either both Pla-Dis or Pol-Dis)
2x Nanite Disruptor Beam Arrays, 1 Polarized Disruptor Beam Array, & 3x Hybrid Plasma-Disruptor Beam Arrays
2 Disruptor Procs
Nanite Cloud Procs
1 Plasma Procs
1 Polaron Procs
(again, dual procs - one Pla-Dis and one Pol-Dis)
The Feds obviously pop too fast to get anywhere with six arrays...so I'm going to try that one again against the Cardassians.
2x Nanite Disruptor Beam Arrays, 1 Polarized Disruptor Beam Array, & 3x Hybrid Plasma-Disruptor Beam Arrays
3 Disruptor Procs
1 Nanite Cloud Procs
2 Plasma Procs
1 Polaron Procs
(again, dual procs...for the Pla-Dis and Pol-Dis)
This was against 2 Keldon, 1 Galor, 1 JHEC, and 4 JHAs.
So uh, yeah...the Nanite thing started to concern me...so I decided to do another test with a friend (er - yeah - a friend). It's probably the test I should have just done in the first place...meh. I let the one guy fire at the other guy for 20 minutes (I needed a smoke or it would have been longer.)
So...firing those six arrays at somebody, I discovered a few things:
1) The Nanite Disruptors do apply a Disruptor proc. But like the Pla-Dis and Pol-Dis, it's a dual proc. Only unlike the Pla-Dis and Pol-Dis, the Nan-Dis only shows a single debuff. This brings up the question of how the proc rolls are handled then, eh?
Is it one 2.5% roll to determine if both are applied or is it two 2.5% rolls to determine if both are applied. Either way there's an issue, but if it's two rolls to apply both, then they will have a higher probability of applying the proc than a non-hybrid.
2) The Nan-Dis Disruptor proc is not "broken" like the Dis, Pla-Dis, and Pol-Dis procs are - there's no "double proc" issue going on.
3) The Nan-Dis and Pla-Dis or Pol-Dis can stack. The debuff from the Nan-Dis and the two debuffs from the Pla-Dis or Pol-Dis will show and the damage resistance will reflect this.
4) All three Disruptor procs can stack - this is visible by viewing the damage resistance dropping, but only two debuffs (well, three) will show. You will only see the Nan-Dis and Pla-Dis or Pol-Dis, but never the Nan-Dis, Pla-Dis, and Pol-Dis.
Nan-Dis drops DR to ~25%
Pla-Dis drops DR to ~7%
Pol-Dis drops DR to ~6%
Nan-Dis + Pla-Dis or Pol-Dis and it's ~1-2%
All three? -10%
edit: Meh, I actually stopped to read the Nanite Disruptor description - it's one proc to do both - so that's fine. It's still going to be an issue on the rolls for the Pla-Dis and Pol-Dis and how they affect uptime if they're not also a single roll.
Course, based on what Hawk said in the other thread...the debuffs should work like the following (if they're supposed to stack):
Energy Resistance DRR 37.5 and 27% DR
DR=(3*(0.25-(75/(150+DRR))^2))
Nan-Dis is -5 DRR
Pla-Dis is -10.6 DRR
Pol-Dis is -11 DRR
A Nan-Dis proc should drop the DR from 27% to...25.1%
37.5 - ((1 - .27) * 5) = 33.85 DR
A Pla-Dis proc should drop the DR from 27% to...22.8% (currently dropping it to ~7%)
37.5 - ((1 - .27) * 10.6) = 29.76
A Pol-Dis proc should drop the DR from 27% to...22.6% (currently dropping it to ~6%)
37.5 - ((1 - .27) * 11) = 29.47
Nan-Dis -> Pla-Dis would drop DR from 25.1% to...20.5% (currently ~1-2%)
33.85 - ((1 - .251) * 10.6) = 25.91
Pla-Dis -> Nan-Dis would drop DR from 22.8% to...20.5% (currently ~1-2%)
29.76 - ((1 - .228) * 5) = 25.9
Nan-Dis+Pla-Dis -> Pol-Dis would drop the DR from 20.5% to...14.6% (currently -10%)
25.91 - ((1 - .205) * 11) = 17.17
Which is a world of difference, and yeah - I expect to see a bunch of complaints from folks on that...meh.
But still, two questions remain:
1) Should they stack?
2) Is there one 2.5% for the Pla-Dis/Pol-Dis to apply both or two 2.5% with the chance to apply both?
If this is true, that is a HUGE nerf. Like, making the disruptor proc the most useless proc.
I was under the impression the "fix" was just to make it always do -10 (or -5 for Nanites, which I was under the impression they said were WAI) and that the math will be fixed so that it is always causing people to take 10% more damage. Not these piddly 3-4% numbers.
Bottom line: Choose your weapons for aesthetics, or if you're commited to power-gaming, choose weapons that get a synergy bonus with a set that suits your playstyle, or even just choose weapons with a cheap tactical console.
This is completely untrue. I have extensively tested proc rates on weapons and found that, depending on your build (AoE build see drastically increased proc rates since the roll is made once per weapon per firing cycle per target.) a beam array boat with 7 beams will average 1 proc every 3-5 seconds when using BFAW and maybe ever 8 to 10 seconds without. This data was averaged over hours of playtime to get as big a sample as possible. A ship with 6 cannon type weapons will average less, but still likely gets a proc off every 15 seconds without scatter volley and 4-5 with.
Now this doesn't mean that they are game breaking (they really arn't) and makes me wish I could get CrtDx3 Antiproton weapons. Likely as is I'll be looking to switch away from disruptors if the fix goes as above.
The guy was pretty explicit that the 10% number he used was not an actual in-game value. He merely used that number as an example because it made what they were doing easier to understand than actually sitting down and hashing out the calcs that go into figuring resistances. The function is fairly simple even if the logic behind using that particular function is fairly arcane, though the fact that it is a function has let to a lot of misconceptions over how exactly resistances work in the game.
And actually if the person stacks enough resist consoles on their ship the disruptor proc's -10 to all resists will result in less than a full percentage point increase in damage done. Provided the disruptor proc is working as intended.
So in the 15 second lifetime of the proc debuff, a beam array will fire 3 times. For that weapon, there's a 92.7% chance that you will NOT get a new proc. Assuming you're firing 8 beam arrays, for any target X, you have a 45.5% chance of refreshing your debuff. So under OPTIMAL conditions, you're debuffing any given target just under half the time. Since the proc rate per target is a function of weapon firing cycle and not shots fired, I have to conclude that any contrary finding are, at best, flawed in their methodology. And it's a conclusion that's supported by my own empirical experience.
The question is really, however, this: Did the debuff truncate the fight more or less than just adding [DMG]x2.
There's this part...
So if you've got a guy with no skills and just a single blue Neut Mk XI giving +17.5 DRR or 14.9% DR and he gets hit by a -10 DRR proc:
The 14.9% DR resists the -10 DRR proc. (1 - .149) * -10 = -8.51 is the DRR debuff applied to the target's DRR of +17.5...giving a new DRR of +8.99 or 8.2% DR.
A 1000 damage shot against the original DR of 14.9% would have done (1 - .149) * 1000 = 851 damage.
A 1000 damage shot against the new DR of 8.99% would now do (1 - 0.0899) * 1000 = 910.1 damage.
Even though the change in DR is 5.91%, the increase in damage is 6.9%.
If the guy had 2x blue Neut Mk XI giving him +35 DRR or 25.7% DR and he got hit by a -10 DRR proc:
Again, the 25.7% DR resists the -10 DRR proc. (1 - .257) * -10 = -7.43 is the DRR debuff applied to the target's DRR of +35...giving a new DRR of +27.57 or 21.5% DR.
A 1000 damage shot against the original DR of 25.7% would have done (1 - .257) * 1000 = 743 damage.
A 1000 damage shot against the new DR of 21.5% would now dow (1 - .215) * 1000) = 785 damage.
The DR change? 4.2% The damage change? 5.6-5.7%
The examples that I used in the earlier post were from my guy that has 27% DR (37.5 DRR).
My finding on this was that no, at least for the armament of my JHDC with 5 beam arrays, a disruptor proc with optimal debuff uptime would not average higher than [DMG]x2. That said, while the debuff IS active, you would be hitting harder than the [DMG]x2 provides, unless firing at an extremely hardened target. I haven't checked for other loadouts however.
The universe has a wonderful sense of humor. The trick is learning how to take a joke.
so is the disruptor proc on the romulan plasma weapons better than the DMGx2 on the advanced fleet plasma or not?
≡/\≡
Yes. Absolutely, no question about it.
That's because in any situation that matters, ESTFs etc., you are not the only one shooting at your target. Think of it this way: DMG is a modifier that affects only the weapon that has it. The disruptor proc applies not only to ALL your weapons, but to every single weapon of every single ship (other players, your pets, their pets etc. etc.) that are shooting your target(s). You'd be crazy to not opt for it.
Sure, but the many shooters, one target scenario also has some limitations to the utility of the disruptor proc too. For one thing, the limit of the effect of the disruptor proc is the hit points of the target. 4 other ships focusing on your target? It's dying in a matter of seconds, disruptor proc or not. There's very, VERY few targets with enough hull to make a noticeable difference in the length of the fight. Second, one disruptor proc from one player can debuff for the whole team. Which means that if someone else planted that debuff, all your ship is doing is overwriting it. Meanwhile the ship armed with advanced fleet weapon is also enjoying the other player's disruptor proc, and doing more damage to boot.
Common misconception. You're not simply overwriting the buff, you're refreshing it. In other words, every player that equips disruptors is increasing the debuff uptime.
See those uptimes calculated upthread? You can extend the result to 16, 32, 40 beams. Or 16 beams, 8 DHCs, and 6 turrets, etc. Every weapon slot, of every player, that is equiped with disruptors / romulan plasma, is increasing the DPS of the whole party by increasing the uptime. From the equation, you can see that the uptime goes asymptotically to 1, as the number of disruptors increases. It never reaches 1, however, so you can always increase the uptime.
This does not even consider the possibility that different debuffs (say plasma and disruptor) may even stack. I don't know that they do, but that would be simply icing the cake.
Also, the whole "resistance resists changes to resistance" worked out perfectly in Hawk's math to show that no matter what the persons resistances were they ended up taking the same amount more damage. (Which is what I meant by the 10%) This makes sense, and should be the way it works. Why should enemy resistance reduce the effectiveness of the disruptor proc?
Not to mention this would make disruptors utterly useless. You'd be better off causing the enemy to lose the defence gained from 25 engine power (not to mention shield and weapon power) than the tiny resistance debuff.
Look, I'm not saying that the disruptor proc is BAD. It's fine. But the low proc rate, marginal utility, and ubiquity of the debuff all bite deeply into its value. To be fair, the added damage from the [DMG] affix isn't stupendous either, but it has the merit of ALWAYS being relevant and useful, and more importantly, the OP has access to them now, as opposed to needing to farm a huge load of reputation to even have the opportunity to obtain MK XII Rom Plasma weaponry.
Now, not everyone will fly a cruiser. But even if everyone were in a science ship (the least number of weapons per ship) you get 5 x 6 = 30 weapons. On the other hand, if people bring carriers, HECs, etc, the number of weapons goes way up. And your disruptor proc, from a SINGLE weapon, will increase the DPS of ALL of them.
The diminishing returns, from the asymptotic approach to full uptime, would have to make the choice of one more disruptor proc at least 30 times less powerful than the increase in DPS from a single DMG modifier, before it would make sense not to equip it.
There was the following...
To try to match that with non-hypothetical flat numbers...
http://sto.gamepedia.com/Damage_resistance
A DR of 40.1% has a DRM/DRR of 70.
100 damage vs. a DR of 40.1% = 59.9
If we apply a resisted Disruptor proc of -10 to that, following the manner in which he describes it: -5.99 would be the proc. The new DRM/DRR would be 64.01. The new DR would be 38.2%.
100 damage vs. a DR of 38.2% = 61.8
An increase of 3% damage. Which looks nothing like his hypothetical example...which makes sense, because the Disruptor proc is a DRR debuff - not a DR debuff.
Even if it were not resisted, say the full -10 took the DRM/DRR from 70 to 60...the new DR would be 36.7%.
100 damage vs. a DR of 36.7% = 63.3
An increase of 5.7% damage.
That amount will always vary (the following are unresisted to show the difference that exists regardless):
100 DRM/DRR (48% DR) 100 damage = 52
-10 DRR (45.7%) 100 damage = 54.3 (+4.4% damage)
75 DRM/DRR (41.7% DR) 100 damage = 58.3
-10 DRR (38.5% DR) 100 damage = 61.5 (+5.5% damage)
50 DRM/DRR (32.8% DR) 100 damage = 67.2
-10 DRR (28.3% DR) 100 damage = 71.7 (+6.7% damage (+6.6% if you don't round...6.69)
25 DRM/DRR (19.9% DR) 100 damage = 80.1
-10 DRR (13.0% DR) 100 damage = 87 (+8.6% damage)
How about the following?
10 DRM/DRR (9.1% DR) 100 damage = 90.9
-10 DRR (0%) 100 damage = 100 (+10% damage)
But again, that's unresisted. Resisted...DRM/DRR drops to 0.91 (0.9% DR)...99.1 damage (+9% damage)...
So was it a case that when he was talking about it resisting - that it was the 4.4, 5.5, 6.7, 8.6 that we saw? Was the example he gave completely hypothetical - as in - well, it had nothing to do with the real math? Perhaps Hawk cloud clarify?
I actually already have access to the Romulan weaponry as well. I figured I was already at the tier 5 reputation stage (been there for a while) and the store unlock itself didn't cost dilithium, so I unlocked them even if I wasn't going to use them immediately and came to the forum to ask the question in my OP.
Just to clarify, I saw the same post by Hawk and used the damage reduction values in the post just above this when checking average damage, I did not use 10% more damage during uptime and I assumed the disruptor proc is not resisted. I also only used positive resistance values, as unless something has changed since last I looked the known formula doesn't quite fit for negative DRR values.
I found that up to about a DRR of 50 (before debuff), the disruptor proc would cause Romulan beam arrays to hit harder than Advanced Fleet [DMG]x2 arrays while the debuff was active. After the DRR of 50 the [DMG]x2 is more effective at all times. I was only using the base item damage for my calculations however, and was not factoring in tactical consoles, weapons power, or skills, although since all of those are multipliers of base damage (I think) it shouldn't change the actual outcome of "which is better," just the scale on which one is greater than the other.
The universe has a wonderful sense of humor. The trick is learning how to take a joke.
Note, I didn't see any extension in my tests (but I only tested for about an hour).
Still, say you have the following...
ProcA1 is about to wear off. ProcA2 lands, which would extend it. But say that an AtS/HE/APO/whatever took place before ProcA2 would have landed. ProcA2, facing current resistance, would be weaker than ProcA1...thus, it would not replace/extend ProcA1.
So that's why I brought up the stuff that Hawk had said...case it wasn't clear why I was pursuing that avenue of the discussion on uptime.
The universe has a wonderful sense of humor. The trick is learning how to take a joke.
Is that true?
It could be that the effect is extended (i.e. re-applied), but is calculated at the new value. (Seems most likely from a programming point of view.)
And surely Proc A did not continue at its prior value when resistance was increased (i.e. HE, etc)
[Kluless][Kold][Steel Heels][Snagtooth]
[Louis Cipher][Outta Gum][Thysa Kymbo][Spanner][Frakk]
[D'Mented][D'Licious]
Joined October 2009. READ BEFORE POSTING
That's where I've asked that he clarify on that. Was his hypothetical example just a figurative example - where it was almost, well...alliterative or allegorical rather than even hypothetical?
Because even without the resist resisting the debuff - it doesn't work out anywhere like what he suggested with the hypothetical example.
This was one of the posts he replied to in the other thread...
...and this was his reply:
Which is why I've gone down the route that I did in my examples.
APD's -47.1 would be resisted by the % DR at the time the debuff was applied. So APD's debuff will not only cause a variable change based on the DRM/DRR of the target, but the actual debuff applied by APD will be variable depending on the % DR when it is applied.
Based on what I quoted above, the debuff value is determined at the point of application and does not change because of other changes, no?
Breaking out my own hypothetical numbers...
ProcA1 is applied when the Target's DR is 50%. A buff wears off of the Target and their DR is now 40%. ProcA1's debuff is still based on that 50%. In this case, if extension does exist - then ProcA2 would have no problem extending...
...but if ProcA1 is applied at 40%, the target buffs to 50%, then ProcA2 wouldn't override the stronger debuff.
But does ProcA2 sit there in limbo, waiting for ProcA1 to expire - and then it applies the remaining duration of ProcA2 would have had? Or does it just...not happen? Tada, the question...
edit: Note, I hate how he avoided the -DR thing...lol. This part from what I quoted myself asking:
That 0 DRM guy would take +1562.7% damage.
The 37.5 DRM guy (27% DR) would take +273.4% damage.
The 73 DRM guy (27% base + APO3) would take +77.8% damage.