Enhancement Proc Changes

Some people specifically build towards the defense soft cap as a deliberate goal, and technically speaking you could argue that for them, the "return" on investment for defense bonuses is linear, and not accelerating. But its a definition of "return" that is sufficiently disconnected from actual in-game value that its a difficult one to honor in the general case.

In your example with Hasten, I think the big motivation is the psychological desire to not have up and down performance. That's qualitatively judged to have a penalty associated with it that is not quantitative. Because actual performance is being ignored to optimize that qualitative situation, that would be an example of one of the corner cases I mentioned. But its also true that another factor comes into play. At the high levels of recharge necessary to get to perma-Hasten, its often the case that you already have enough to make a full attack chain. A meta-concern now factors in. Recharge may be continuing to improve each individual attack, but its no longer improving the chain. At that point, the *only* thing you're getting from that extra recharge is reducing that on-paper downtime of Hasten, and once that reaches a sufficiently low level for the up/down issue to no longer be a concern (the moment its low enough to "ride out") the motivation for improving recharge disappears, even if there are still quantitative benefits.

In other words, its often the case that "return on investment" becomes a non-issue in terms of the raw numbers, and what's being looked at are more holistic issues of build management. In that situation, its too complex to say what the actual "return on investment" is for different kinds of changes except when dealing with very specific changes to specific builds that can't always be extrapolated.

Which changed when it became "Will Proc 4 times a minute" vs "20% Chance to fire off"

Nothing random about "Will Proc 4 times a minute"

That makes as much sense as saying there's nothing random about 20% of your attacks triggering the proc.

The PPM is an aggregate rate, not a schedule.

Originally Posted by Arcanaville Softcapping an Invuln is fantastic. Softcapping a Willpower is amazing. Softcapping SR is kissing your sister.

I'd like to see the proc floor raised a bit as well. Synapse has stated that there are very few powers that actually hit that proc floor, but that's because the proc floor is so low.

Current floor is 1.5(Current PPM*1.25)+5

5% proc (1 PPM) ----------------> 6.875% -----------> (137% of original)
15% proc (2.5 PPM) ------------>9.6875% ----------> (64.6% of original)
20% proc (3 PPM) --------------> 10.625% ----------> (53.125% of original)
25% purple proc (3.5 PPM) ----> 11.5625% --------> (46.25% of original)
33% purple proc (4.5 PPM) ----> 13.4375% --------> (40.72% of original)

The issue I see here is that the minimum proc chance is much better for low proc IOs but strictly worse for higher proc chance IOs such as purples.

What I feel it should be is 3(Current PPM*1.25)+0

5% proc (1 PPM) ----------------> 3.75% ---------> 75% of original
15% proc (2.5 PPM) ------------> 9.375% ------> 62.5% of original
20% proc (3 PPM) --------------> 11.25% -------> 56.25% of original
25% purple proc (3.5 PPM) ----> 13.125 --------> 52.5% of original
33% purple proc (4.5 PPM) ----> 16.875---------> 51.136% of original

TBH, I'd prefer that modifier to be 3.5 rather than 3, or heck even 4, however I think 3 is the best way to get it looked at seriously. The main thing is that minimum proc rate is more even between PPM levels and never dips below 50% of the original proc chance. It would be slightly better for procs that were originally 20%/3 PPM or higher, about even for 15%/2.5 PPM, and slightly worse for 5%/1 PPM, though I think the last is well justified as I don't think the minimum should really be higher than the original proc chance.


If I made any math mistakes please let me know.


edit: Just for gits and shiggles,
3.5(Current PPM*1.25)+0 would look like this:

5% proc (1 PPM) ----------------> 4.375% -----------> 87.5% of original
15% proc (2.5 PPM) ------------> 10.9375% --------> 72.92% of original
20% proc (3 PPM) --------------> 13.125% ----------> 65.625% of original
25% purple proc (3.5 PPM) ----> 15.3125% --------> 61.25% of original
33% purple proc (4.5 PPM) ----> 19.6875% ---------> 59.66% of original

These would be my preffered numbers.



edit2: on the contrary to this, just making the minimum be 66% of the original proc value would work just as well or better.

Agreed in the specific example of Hasten, for most builds, because as you point out the performance difference once downtime is low starts to become meaningless, but as you said yourself,

This is very much a topic where the devil is in the details. Just a few other examples to make that point are Phantom Army, Fade, and the attack chain for Titan Weapons.

In the former two, the difference between 10 seconds of downtime or 0 is not just qualitative, as it's not hard to imagine a scenario in which the downtime on one of those abilities could pretty easily result in death.

In the latter, If I recall correctly the necessary recharge to sustain the optimal attack chain is something like 300% in Follow Through and near that in Arc of Destruction. Having a recharge fluctuation during that attack chain would impact DPS (Momentum mechanic aside, but that has the potential to make it much worse depending on timing).

As I said before, the absolute value of reduction is definitely not the only consideration, and you're probably right that it's not the most important consideration, but I don't feel it's necessarily as disconnected from the game mechanics as you seem to.

I24 Proc Chance = (Enhanced Recharge + Activation Time) * (Current PPM * 1.25) / 60*(1 + .75*(.15*Radius - 0.011*Radius*(360-Arc)/30))
Single Target Radius = 0. AoE Non-Cone Arc = 360.

My bad. I admit I didn't read the entire humongous thread before replying. Didn't realize Synapse had changed the proposal. Good God he has a lot of post in this thread... *goes back to reading*

It's achieving that on average. 3 PPM in a power with a 10s cycle time has a 50% chance to go off on every activation. In a power with a 5s cycle time, the same proc has a 25% chance to go off on every activation. Those are still random. They just work out to 3 activations per minute on average if you always fire those powers as fast as possible.

And it doesn't say it "will activate X times per minute". Here's the actual text present on PPM procs.

Emphasis mine.

Programmed Random Occurance is a backronym. Proc is short for Procedure, and a good way to see this in action, in game, is Proc120s. They are 100% chance procs, that happen and persist for 120s.

The Procedure is figuring out if the effect should take place, and the design intent is actually conflicting. PPMs are a procedure that attempt to ensure different cycle times get a comprable number of effects per minute. The original procedure's intent, however, is less concerned with being comprable and more focused on static random.

100% rates are not against the nature of procs, but instead against the original design intent which conflicts with the new proc mechanic intent.

Actually, Synapse pretty much directly contradicted the belief that Miracle/Numi/Celerity are the same thing as "X% chance of" procs, earlier in this thread.

"You got to dig it to dig it, you dig?"
Thelonious Monk

I'm not saying they're the same thing. I'm saying "Proc" doesn't necessitate random, the design of the type of proc does. Saying "It's a proc, and procs are random" is a falacy. Proc is simply a way of determining if an effect should take place, that could be random or some other qualifying characteristic.

This is important, because with the two types of designs we're looking at, we have a chance for occurance and a chance for occurance normalized for averaging a number of effects per minute.

If procs don't have to be random by virtue of being procs then it's just a conflict between the two design intents, needs to be random vs. targeted number of effects per minute, and we may be able to remove the cap.

This would be a credible theory, if the mechanics of such enhancements was procedural. They are not. In fact, Numina's Convalesence +Regen/+Recovery for example is what you might call a "broken enhancement." Its a power that is flagged as an enhancement, with effects that are flagged to affect the entire player.

It has a 100% chance for going off because by design *everything* has a 100% chance to go off by default. There's no such thing as an effect with no chance associated with it. Power Bolt deals 100% chance for smashing damage and 100% chance for energy damage. But we don't consider them procs.

If it helps, I believe Zombie Man is badly indifferent.

PPM procs and "flat rate" procs are both random. The only difference is that the PPM proc has code that varies the value it compares to the number it gets back from the RNG to to check for activation, where the "flat rate" proc always checks the RNG against a static value. The PPM procs use this dynamic control to set the activation threshold in a way that targets a given PPM activation rate.

In my opinion, targeting a PPM rate and simultaneously setting a ceiling are, in fact, in conceptual conflict. However, that's being done because there is actually an external design constraint being imposed - that random things should not be allowed to achieve a consistent 100% chance to happen, because then they cease to be random. PPM procs are random things, so their chance to activate should not be allowed to reach 100%.

Edit: To use Arcanaville's post as a springboard for further clarity, what most of us refer to as "procs" conceptually act like this.

On Activation of Slotted Power:

• Get number from RNG
• Compare number to activation threshold
• If RNG value was > threshold
  • Activate additional effect
• Else
  • Do nothing

In contrast, things that are not procs, like Miracle unique pieces, conceptually act like this.

On Activation of Slotted Power:

• Activate additional effect

I believe that what Synapse is saying is that things that act like my first pseudo code should not be allowed to achieve a threshold value that reaches 100%, meaning they always get an RNG number and then always pass the threshold test. If they perform that test, they should have a limit on the threshold so that it's always possible to fail the test.

Nope. This is some crazy backronym people made up. It's short for "procedure". If you have a special procedure that executes under some circumstance, that's a "proc".

Arguably, with a 20% proc rate, the proc always happens, it's just that what always happens is a 1-in-5 chance of doing damage. :P

So nice of you to say since you're so indifferently good.


Anyway, terminology:

The backronym of Programed Random Occurrence comes from the WoW community. There is no evidence of that phrase existing before Procs were introduced in WoW. Meanwhile, Positron, who introduced us to 'procs' was used to that term from Everquest days which predated WoW and he defined it as a procedure.

But that term and others like 'Uniques' got applied to all the Special Single-acting IOs... incorrectly... or correctly... depending on how Positron and Castle defined 'procedure.'

What we players have determined is that there are three categories of Special Single-acting IOs which we (and be 'we', I mean 'me' since I cleaned up the terminological chaos in the Wiki and mostly everyone has accepted it ) have called them:

1. Global IOs: Act like Global Set Bonuses (always on). Obey the -3 level minimum cut off rule.

2. Procs: Chance to fire off when the power is used. Doesn't obey the -3 level minimum cut off rule.

3. Proc120s: Just like Procs, except they last 120 seconds in clicks and pulse 10 seconds in toggles/autos, and their 'chance' to go off is 100%.

Whether that third category should be called a proc or not has been debated. But it's a widely accepted name to get a handle on what it does in apparent effect regardless of actual mechanics.



As to my opinions for my one fan:

• I believe that if just one single 'proc' -- whether legacy or PPM -- is being relied upon by an AT to make a *significant* jump in 'power' or to make the AT 'playable'; then it is Overpowered. That may be the case with some of the new PPMs and if a cap is needed, cap 'em. A proc should be as good or slightly better than a Level 50 Damage IO in effect. If it's several orders better... that's not good.

• I also believe too many of the Legacy procs are underpowered and worthless and need a buff.

• I have guarded optimism this will all be worked out satisfactorily by I24.

• I also believe the powers team should have heeded previous warnings and not brought this to Live so quickly.

• I also believe that the level ranges of Enhancement Sets needed more of the Powers Team attention than the procs ever did. Level 10-20 Sets? Useless and ridiculous.

• Also, the Powers Team should have focused on new Sets for categories that are short on options.

But what does someone so badly indifferent know?

This is my understanding as well. It comes from the name of the field in the old MUD code (DikuMUD, I think), where you could attach a "procedure" to an item that would run when the item was used or similar. The procedure always ran when the item was used, but typically it "rolled a die" and determined a random chance for something to happen.

It isn't really important, but I kinda hate it when people muck up perfectly easy to understand history.

Awesome. Make that seven years for me, and sign me up.

(I can count the number of times I've disagreed with Arcana on zero hands. At least, once I understood what she was saying... freaking complicated sometimes. )

Right, and the EQ guys where old MUD programmers, and they used a lot of the same ideas and terms.

I'm hopeful that some of those will get a second glance when the devs convert them all to PPM. 10% chance for 5% heal in a single target power is a little low. 2% chance for disorient is just plain ridiculous.

Of course, the king of what the heck isn't a chance proc: its Gift of the Ancients Def/+7.5% run speed, each of which adds 1.07 mph to your running speed. Fortunately, I convinced Castle during I9 beta to change it from being Unique. You're welcome.

For allowing me to waste 4 extra slots? Yeah, thanks for that!

It should be noted I do have 2 characters with 5 of those slotted. I still think EVERY +speed bonus in IOs should be increased by 50%.

Stone Brutes aren't popular anymore, therefore that proc isn't popular anymore, whether it's unique or useful or not isn't really relevant because the only thing that would actually invest in them is a powerset that is about as popular as Energy Melee.

Yeesh. You know that swift is slottable, right?

If you're chasing IOs for speed, you have to also deal with what kind of speed: run, jump, or fly -- and there's precious little of each. My suggestion, pitched several times, is to change all the extra speed IOs to Universal Extra Speed and then buff them a bit.



Every extra slot in Swift past the first is relatively as worthless as a +Speed IO. Allowing players to put extra slots in Swift is.....

<get your shot glasses out>

... a slap in the face!!


Oopah!

I get 2 mph out of that 2nd slot in swift! That is huge. The tier 4 Musculature allowed me to get away with only 3 slots in Sprint too!