Damage Procs and PPM tweaks or changes - Suggestion thread

[Caulderone]

For info on the current PPM setup, Bopper's very detailed and informative thread is linked at the bottom of this post.

 

It is fairly obvious from a balance perspective that the PPMs changes instituted by the original Devs in I24 have lead to many unintended interactions that weren't discovered by many folks prior to shutdown.  Example:  Turning a long recharge debuff power into a Nuke.  Here is Captain Powerhouse's last comment on that:

" This is more of an issue with PPM procs, though, something that needs addressing sooner or later. "

 

 

So, I started thinking about what I would change were I doing the tweaks.  There are lots of options.  The old percentage based proc system favored fast recharging powers and had minimal benefit to long recharge powers.  The new system limits the benefits to fast recharging powers and quite obviously way overproduces on long recharging powers.

 

So, I think we should start brainstorming ideas that we can present to CP for that "addressing sooner or later" time.  Here are my initial thoughts.

 

1.  From Bopper's thread:

Maximum Probability to Proc:

MaxProb = 0.90     

*Note* Max probability is 90%

So, this is the problem child in long recharge powers.  Lowering the max probability is an easy change for reducing the overproduction of damage procs in long recharge powers.  My initial thought, based on no math as of yet, is to set the cap at 50%.

 

2.  From Bopper's thread:

Minimum Probability to Proc:

MinProb = 0.05 + 0.015 * PPM

*Note* Min probability is 5% + PPM x 1.5%

On this one, fast animating and recharging powers like Shadow Punch or Dark Blast get varying utility from damage procs if you slot any recharge at all, but do ok if you don't end up slotting recharge.  Example:  3.5ppm gives ~29% proc chance at 0 recharge slotting and ~18% at 0.8 slotted recharge.  While I feel like I would like the minimum to be a bit higher, this generally works out as fairly reasonable to me, as it isn't much different from the old base 20% proc chance and can be better with no recharge slotting.

 

So then, my initial proposal is a very simple one.

---->>>  Lower the maximum probability built into the existing formula to bring very long recharging powers back down to reasonable levels while still letting them leverage damage procs better than the prior system did.  I think that would be all that would be needed.

 

Please chime in with your thoughts and suggestions!

 

 

 

Here's that Bopper thread:

 

[Bopper]

Ugh...time to fall on my sword.

 

Two options I have been thinking about, which is a huge neuter, but probably goes back to the original intention of PPM.

 

1) Go back to using total recharge, not enhanced recharge. The main exploits come from getting large global recharge without slotting recharge. So this alleviates that. But if I could suggest a compromise, use only set bonuses and enhancements for the MRT, while allowing outside global buffs (like Hasten and Speed Boost) to still not impact proc probability. It will still neuter current proc performance without taking some fun out of trying to build around some optimization methods.

 

2) Have all Proc Enhancements use a duration that is equal to 60/PPM, instead of the default 10 seconds. Then have each Enhancement carry a flag to determine whether or not it had proc'd. If it had proc'd, it can't proc again until the duration of the enhancement wears off and its effects refresh (thus resetting the flag).

 

Some problems with the second suggestion is with very long AoE attacks that have an attack rate that is longer than the duration of the enhancement. This suggestion would limit it to only the 1 proc, thus underperforming the advertised PPM. The reason why it's a flaw is because hitting multiple targets (and procing on multiples) could help bring the average proc rate back in line with its PPM. So some extra thought needs to go into that idea.

[Auroxis]

From what I gathered seeing what Captain Powerhouse wrote about this topic, his direction will probably be more towards how damage procs interact with class and/or power damage multipliers.

 

Edited March 10, 2020 by Auroxis

[Caulderone]

7 minutes ago, Auroxis said:

From what I gathered seeing what Captain Powerhouse wrote about this topic, is that his direction will probably be more towards how damage procs interact with class and/or power damage multipliers.

Hmm..  Interesting idea.  I hadn't thought of tying the proc damage to the archetype scalars.  I'm not sure that alone would fix the issue.  I feel like it might be a whole lot more work, too, though.

[Bopper]

3 minutes ago, Caulderone said:

Hmm..  Interesting idea.  I hadn't thought of tying the proc damage to the archetype scalars.  I'm not sure that alone would fix the issue.  I feel like it might be a whole lot more work, too, though.

It would shift the balance from low damage ATs using procs to close the damage gaps to having higher damage ATs using procs to widen the damage gap. It would trade one problem for a different problem (maybe a worse problem). But I am purely speculating. Without a formula, there's no way to truly analyze what the impact would be.

[Auroxis]

2 minutes ago, Bopper said:

It would shift the balance from low damage ATs using procs to close the damage gaps to having higher damage ATs using procs to widen the damage gap. It would trade one problem for a different problem (maybe a worse problem). But I am purely speculating. Without a formula, there's no way to truly analyze what the impact would be.

To be fair, ANY nerf to procs would have the same effect in the grand scheme of things.

 

But yes, it depends on the formula.

Edited March 10, 2020 by Auroxis

[Bopper]

Just now, Auroxis said:

To be fair, ANY nerf to procs would have the same effect.

Absolutely. And I wasn't trying to shoot the messenger, I know you're just interpreting what he said. My response was strictly in relation to what I think that implementation would do

[Bopper]

As for a 3rd proposal, it is similar to my 2and proposal. But instead of the flag only allowing the possibility to proc on a single target then locking out afterwards (which might be hard to mechanized anyways), have the flag turn off for future uses of the power if the enhancement does proc atleast once. This would allow for AoEs to still have an opportunity to proc on each target it hits.

 

Then, give the proc a default 90% probability to proc, and have that proc probability use the original Area Factor to reduce the probability for AoE attacks.

 

In summary:

Proc Enhancement duration: 60s/PPM

Proc Flag defaulted/refreshes to True (can fire)

Proc Probability = 90% / [1 + Rad*(11*Arc+540)/30,000]

If a target gets hit by Proc, set Proc Flag to False (can't fire) for remainder of Enhancement duration.

 

With that implementation, you will have a proc rate that more closely matches PPM. And likely nobody would use Procs unless it's for the set bonuses.

 

 

[Caulderone]

Been trying to figure out a way to reasonably measure my 50% proc chance cap to see how much of a nerf it would be.

 

3.5PPM proc does 71.75 damage

71.75 * 0.9 = 64.575 avg dmg

71.75 * 0.5 = 35.875 avg dmg

which means the procs would do ~55.55% of the current damage for things at 90% chance cap

Edited March 10, 2020 by Caulderone

[Bopper]

5 minutes ago, Caulderone said:

Been trying to figure out a way to reasonably measure my 50% proc chance cap to see how much of a nerf it would be.

 

3.5PPM proc does 71.75 damage

71.75 * 0.9 = 64.575 avg dmg

71.75 * 0.5 = 35.875 avg dmg

which means the procs would do ~55.55% of the current damage for things at 90% chance cap

And the formula for achieving capped procprobability becomes 

 

Allowed Recharge = [BaseRecharge/(30*AreaMod/PPM - Cast) - 1] * 100%

 

Also, a 3.5 PPM would require a MRT of 8.58s*AreaMod. 

 

So I project the effects of lowering the max proc probability to 50% would result in attack chains electing to use faster attacks, such as a ST attack with base 8s with cast 1s would give you 50% proc probability, while global recharge likely knocks down the cooldown to something like 3s, making it easy to make frequent use in an attack chain.

 

The question would become, is it worth the average 35.875 damage. That really depends on the enhancement options. Fast attacks tend to be weaker, so probably with use procs in them. Same with AoE attacks. But at least they wont be turned into nukes.

 

All that being said, here is my simple rule (principle?) for ensuring the definition of PPM is maintained. If you have an attack hit rate that is faster than the proc rate, you should be able to achieve a proc rate of PPM. In our current system, we don't technically satisfy that because there are conditions that would result in the 90% cap to not allow for achieving a proc rate of PPM, but atleast it ensures us we will atleast always hit a proc rate of at least 0.9*PPM.

 

By lowering the cap, you will introduce a fairly wide range of conditions that would result in hitting a target at a rate greater than PPM, however the cap prevents you from achieving a proc rate of PPM. Instead you may result in performance as low as 0.5*PPM.

 

With that much disparity, we would probably have to get away from calling it PPM altogether.

[Leo_G]

3 hours ago, Bopper said:

1) Go back to using total recharge, not enhanced recharge. The main exploits come from getting large global recharge without slotting recharge. So this alleviates that. But if I could suggest a compromise, use only set bonuses and enhancements for the MRT, while allowing outside global buffs (like Hasten and Speed Boost) to still not impact proc probability. It will still neuter current proc performance without taking some fun out of trying to build around some optimization methods.

I think this might be the more straight forward solution.

 

I don't have a problem with an AoE debuff power with no damage being turned into a nuke if said nuke was only usable every 120sec or so instead of using global recharge + buffs to bring that nuke down to 50sec or less.

 

Another solution, what if proc chance was slightly tied to matching sets?  So slotting a proc and 1 other IO from the set doubles its proc chance, 2 triples it (max)?  So there's the option of having 4-6 procs all with small chances or use 2 procs + 2 of their respective sets for more consistent activation thus limiting how much proc damage would occur?

Edited March 10, 2020 by Leo_G

[aethereal]

I don't feel like I understand the design goal for procs.

 

What are they supposed to do?  Is the idea to allow people to improve average performance at the cost of unreliability?  Is that...  actually something that anyone wants?  Certainly, in practice people seem pretty dedicated to maximizing reliability.  Is it just supposed to be a complicated system that allows people who have additional systems mastery to improve their performance by demonstrating mastery over fairly complicated game mechanics?  Is it supposed to make powers more diverse (if so, we have too many damage procs), and the unreliability is beside the point?

 

I don't really get it.

[Caulderone]

20 minutes ago, Bopper said:

So I project the effects of lowering the max proc probability to 50% would result in attack chains electing to use faster attacks, such as a ST attack with base 8s with cast 1s would give you 50% proc probability, while global recharge likely knocks down the cooldown to something like 3s, making it easy to make frequent use in an attack chain.

So, basically, I just shifted the goalposts, but it did alleviate the long recharge powers = nukes.

 

17 minutes ago, Leo_G said:

I think this might be the more straight forward solution.

 

I don't have a problem with an AoE debuff power with no damage being turned into a nuke if said nuke was only usable every 120sec or so instead of using global recharge + buffs to bring that nuke down to 50sec or less.

 

Another solution, what if proc chance was slightly tied to matching sets?  So slotting a proc and 1 other power from the set doubles its proc chance, 3 triples it (max)?  So there's the option of having 4-6 procs all with small chances or use 2 procs + 2 of their respective sets for more consistent activation thus limiting how much proc damage would occur?

Yeah, using total recharge does seem to look better here.  Although, then I would want serious consideration to go to raising the minimum chance.  Otherwise, the fast recharging powers are basically going to be locked at minimum proc chance permanently.

 

Tying the proc chance to having other IOs from a set slotted is interesting.  Sounds difficult to enact, though.

Edited March 10, 2020 by Caulderone

[Grouchybeast]

Cap proc damage at a percentage of the base damage of the power.

[Leo_G]

Also, can we separate debuff/control proc chances from damage proc chances (and buffs)?  There are some effects that I'd like to use but the costs to slotting a -rech or knockback proc that rarely triggers feels a bit underwhelming.

[Bopper]

5 minutes ago, Grouchybeast said:

Cap proc damage at a percentage of the base damage of the power.

What about powers that dont do damage, or procs that don't do damage?

[Grouchybeast]

11 minutes ago, Bopper said:

What about powers that dont do damage, or procs that don't do damage?

Good point.

 

Maybe just simply limit damage procs to one per power?

Edited March 11, 2020 by Grouchybeast

[nzer]

50 minutes ago, aethereal said:

I don't feel like I understand the design goal for procs.

 

What are they supposed to do?  Is the idea to allow people to improve average performance at the cost of unreliability?  Is that...  actually something that anyone wants?  Certainly, in practice people seem pretty dedicated to maximizing reliability.  Is it just supposed to be a complicated system that allows people who have additional systems mastery to improve their performance by demonstrating mastery over fairly complicated game mechanics?  Is it supposed to make powers more diverse (if so, we have too many damage procs), and the unreliability is beside the point?

 

I don't really get it.

Presumably the second option, but I don't see it as rewarding mastery over obtuse systems as much as providing build tension for minmaxers to chew on. In that I think it's pretty successful, while having the side benefit of allowing support ATs to build for some damage if they want to. The problem is that offers so much benefit to abilities that aren't meant to be relied on for damage that they can end up superceding abilities that are.

[Bopper]

3 minutes ago, Grouchybeast said:

Good point.

 

Maybe just simply limit damage procs to one per power?

That feels like an unnecessary restriction. The beauty with procs is that it offered playstyles that weren't traditional. You were able to create your own style and it could be fun and creative. Restricting players to conform to rules on how they can slot essentially suggests there's only one right way to play. This type of response makes me fear that we will go back to builds strictly chasing the best set bonuses: purples, winters, ATOs, and plug the rest with recharge and defense. Itll feel like builds and powerset combinations will become cookie cutter again.

 

All that being said, I wouldn't be surprised if they approach this similarly to enhancement diversification, where perhaps for each like-proc that fires from the same power will have a diminished effect. So if there are 6 damage procs (1 purple), if the purple fires, it has top priority and has 100% effect. If a 2nd proc fires, it has 80% effect. A 3rd gets 60%. A 4th gets 40%. A 5th gets 20%. A 6th gets 10%.

[Leo_G]

Another off-the-wall idea: What if procs themselves were really low damage BUT benefit from damage slotting (not buffs...but maybe set bonus +damage?)?

 

My brain train was aiming for making fewer procs (1 or 2) more optimal so you could have a damage proc that produced more damage getting you a kind of sense of customizing your damage types.  Instead of slotting 3+ procs with built to go off at optimal rates for maximized damage to just 1 proc that does the equivalent of 2 procs of damage.  Powers that don't do damage (thus can't be slotted for damage) would likely benefit more from multiple procs instead but would present a diminished effect because they can't enhance the procs' damage.

 

Although now that I think about it, such a point likely keeps the status quo the exact same but with different slotting strategies.

[Sarrate]

I'm not an expert wrt to procs (your guide was fantastic, Bopper), and what it really sounds like we're trying to solve multiple, competing problems at the same time.

 

1) When using a flat rate, fast recharging powers were disproportionately powerful (eg: Neutrino Bolt) while slower powers were left out.

2) Using PPM and being affected by recharge (enhancements, set bonuses, etc, or some combination of the two) leads to fast recharging powers being overly penalized. It also leads to really weird build behavior where people might avoid recharge to maximize procs. It "feels" wrong.

3) Long recharge powers basically guarantee procs and thus can be transformed into mini nukes.

 

I feel like making the proc rate more sensitive to recharge exasperates #2, and I think that just flat out making them only proc X times per minute basically just makes them a timed proc rather than a random one. (I also can't help but wonder if it would make the alpha strike problem worse, as all your procs would be refreshed at the start of the fight, fire, then potentially get refreshed for the next spawn.)

 

 

I was thinking @Caulderone 's was on the right track of having a higher minimum and lower maximum to keep thing random. It could lead to #1 potentially being problematic, but that's probably less of a balance problem than inadvertently creating nukes. An option that could be consider if fast attack + recharge + higher proc minimum was too much is scale the proc's damage downward by some amount based on how much a power is beating the minimum proc rate. Of course, that makes procs less exciting when they happen - as they'd be less impactful.

 

 

The real question is what is the priority of the problems with procs? Once you know that, then you can fashion a solution around those priorities.

[MunkiLord]

Any massive change to procs would mean I'd be doing several respecs and would be quite a time investment. Personally, I am 100% opposed to a change that impactful until we get a respec system that doesn't suck. A better respec system should absolutely be a higher priority than fixing procs. It needs to come first.

Edited March 11, 2020 by MunkiLord
typo

[oedipus_tex]

Thanks for bringing this up. I don't think this is an easy fix. There are some good ideas above however.

 

 

The main reason the proc system favors long recharge powers is that a procs per minute system does not account for animation time. If a power recharges in 4 seconds and takes 1 second to animate I currently have to expend 15 seconds of animation time, casting it over and over, to produce the expected output. Where a power with a 1 second animation, 30 second recharge I might only need to spend 2 seconds for the same payload.

 

Fixing this is a math question somewhat beyond my abilities, and also partly a philosophical question about how much "certainty" is worth as the chance edges toward 100%. But thats the crux of it.

[Apparition]

30 minutes ago, MunkiLord said:

Any massive change to procs would mean I'd be doing several respecs and would be quite a time investment. Personally, I am 100% opposed to a change that impactful until we get a respec system that doesn't suck. A better respec system should absolutely be a higher priority than fixing procs. I needs to come first.

 

Not really.  The biggest problem with damage procs IMO is that they completely invalidate the Corruptor AT.  Before sunset, even in Issue 23, you saw a fair share of Corruptors around.  Now?  You see one in a blue moon.  Why?  Because most Defenders can now out buff, out debuff, and out damage most Corruptors.  IMO, fixing that extreme disparity is a bigger priority than a respec system.

[nzer]

40 minutes ago, Sarrate said:

I'm not an expert wrt to procs (your guide was fantastic, Bopper), and what it really sounds like we're trying to solve multiple, competing problems at the same time.

 

1) When using a flat rate, fast recharging powers were disproportionately powerful (eg: Neutrino Bolt) while slower powers were left out.

2) Using PPM and being affected by recharge (enhancements, set bonuses, etc, or some combination of the two) leads to fast recharging powers being overly penalized. It also leads to really weird build behavior where people might avoid recharge to maximize procs. It "feels" wrong.

3) Long recharge powers basically guarantee procs and thus can be transformed into mini nukes.

We can solve all three of these at the same time, I think, by applying a weighting to recharge time in the PPM formula to skew the proc chance more in favor of short recharge powers. This will make it harder to turn long recharge powers into nukes, compensate for the cost of animation time in short recharge powers @oedipus_tex mentioned, and lower the damage loss of additional recharge so people aren't avoiding slotting recharge because it will hurt their proc damage.

 

The downsides are that it turns procs into even more of math problem than they already are, and renders the actual PPM text meaningless, as the number of procs per minute would depend on the recharge time on the power.

Edited March 11, 2020 by nzer