Theory: Proc Rate coded by proc, not weapon based

[Wompem Ondahed]

Through the miracle of caffiene and lack of sleep a random thought occured to me that I have yet to read. Most people say procs are weapon based/time based but I have a theory with 2 different annecdotal evidences (sp?).

It is taken for granted that weapons are set at around 2 procs per minute. There is seat of the pants evidence/parses that such weapons as Bloodfrenzy and the cleric summoned hammers are set higher. What many do not think about, if a weapon alone determines proc rate then people augmenting their non-proc weapons would get 0 procs. In the creation of weapons Verant didn't say " hey, we should give rusty weapons the chance to proc rate of two times per minute, just in case someday we add a proc onto them!"

I would tend to think procs are coded to go off a certain %, and some weapons may have modifiers within. Say for instance Serverside Bloodfrenzy has ~+25% proc coded, cleric hammer has ~+50%, whatever ( theoretical numbers of course).

More theoretical evidence would be proc buffs such as Call of Sky etc. These buffs are temporary, and therfore if weapons solely determined proc rates then they would never go off if you had a non-proc weapon equiped.

What this means: I have no clue, take it as you will. I merely set this before you as something to ponder, such as why BF procs more than BoC etc etc.

[Gerald]

I don't think this is the case because then you'd see multiple spells in the spell data file. There is one instance of Enraging Blow and the proc is on 4 different weapons. Bloodfrenzy procs more than the others. What would this lead you to believe?

It's a guarantee that augmented procs are coded seperately from weapons. It probably works like a weapon now has 2 proc slots. A weapon's primary proc slot is coded to whatever the developers set it at (usually 2 or so). Secondary procs are most likely all set at 2 procs per minute.

[Tiejinnie]

Just ask any Cleric who plays with their Summoned Hammer and they'll tell you that "PROC" can be coded into the weapon.

Certain weapons are coded to "proc" more often than others.

Don't believe so?

Check out these Cleric-summoned Hammers:

Level 58:
Hammer of Divinity

Level 60:
Hammer of Souls

Level 63:
Hammer of Damnation

All 3 are summoned hammers. When it was first released, it was announced that the Hammer of Damnation was coded so that it would proc more often than the former 2.

I also play a 57 Cleric (and my wife's a 65 Cleric). When these spells were released, we realised that they could actually control the actual Proc-Rate of an item to a certain degree (if we factor in RNG).

Personally, I believe that the rate of proc is coded according to the type of proc, and certain weapon-models have a modifier to it, to increase the proc-rate (ie. BloodFrenzy).

[Raiddinn]

Whats wrong with the idea that all procs are coded to go off 2-3 times per minute and for certain weapons it just checks for the same proc more than once per swing? Like for cleric summoned hammer maybe there is a loop to check for proc like 10 times a swing?

It would take all of 2 lines of code to do it like that, and most likely more to do it any other way.

Impressario Raiddinn the Beatdropper
65/243 Maestro Rallos Zek

[Wompem Ondahed]

I think Raiddinn's idea is closest to my own. I suppose it is an abstact idea that does not matter much anyhow, but I want to make sure people understand what I am saying.

Basically i tend to think that most procs are coded a certain %, and that weapons could in fact have a % proc modifier. It would make sense, just as high Dex has shown to have a small modifer when considering proc rates. Indeed it would be simple to program:

Standard_Proc_01 = 2 procs per minute ( or whatever value ).
Split Blade of Destruction proc mod = none ( obviously why would they re-code a normally non-proccing weapon?)
Blade of Carnage proc mod = none (standard modifier of none)
Blood Frenzy proc mod = +25%
Cleric Hammer proc mod = +75%

Anyhow, again this is very obscure and totally open to interpratation. It all stems from curiousity as to why a weapon with no proc would have ever been coded with the future idea that it could proc, so I was thinking procs are set, with weapon containing the mod %.

[EmiliaEQ]

Lets get a few things set out straight so we all understand :

1) The average proc rate is 2PPM
2) This proc rate is with 305 dex
3) Dex affects proc rate (need 4 5 parses to determine how)
4) Some weapons proc more than others
5) Haste doesnt affect overall PPM
6) Slows/Haste Increase/Decrease the overall Proc Per Swing

Actually what everyone fails to realise its the formula VI uses
is based on a PROC PER ATTACK basis :
(Making a PPM formula based on Time and not Attack is a nightmare)

Proc Rate = CEM * 2/(600/DLY) * (305/305) * 1/(100+HST) /AAC

DLY = Weapon delay
CEM = Combat Effects Mod = (1+CE/100+WCE/100)
CE = Combat Effects
WCE = Weapon Combat Effects
HST = Haste factor (Haste/Slows)
AAC = Average Attack Coefficient (Avg attacks per swing)

How is the AAC calculated ? Simple, how many attacks per swing
is the person in questions supposed to land (see skills)
Single attack = 100%
Dual Attack = 80%
Triple Attack = 10%
AAC for that person is 1.9
(I know its less that 1.9 but i'll use 1.9 caus its simpler)


Case 1) 20 delay weapon 0% haste

2/(600/20) * 305/305 * 100/(100+0) / 1.9 = 3.5%

You will have a 3.5% chance of proc per attack.

20 delay = 30 swings per minute
1.9 attacks per swing = 30 x 1.9 = 57 attacks per minute
57 attacks at 3.5% proc rate = 2 PPM

Case 2) 20 delay weapon 100% haste

2/(600/20) * 305/305 * 100/(100+100) / 1.9 = 1.75%

You will have a 1.75% chance of proc per attack.

10 delay = 60 swings per minute
1.9 attacks per swing = 60 x 1.9 = 114 attacks per minute
114 attacks at 1.75% proc rate = 2 PPM


Case 3) 60 delay weapon 50% slowed

2/(600/60) * 305/305 * 100/(100-50) / 1.9 = 21%

You will have a 21 % chance of proc per attack.

120 delay = 5 swings per min
1.9 attacks per swing = 5 x 1.9 = 9.5 attacks per minute
9.5 attacks at 21 % proc rate = 2 PPM



PS 1 : 50% slow will double your weapons delay, while 100% haste will half it.
Ps 2 : Need some 2PPM dex parses to tweak the proc to dex ratio

I believe that some weapons (BF DBoW BBoE) have innate self only combat effects.
Thus explaining their higher PPM.
This goes very well with the fact that CE affects PPM

[Ladas Sindarin]

Nice examples... except you used the same AAC, which is derived off a formula related to % of number of attacks per swing... and as far as I know, clerics don't get triple attack... or double attack... so wouldn't a clerics AAC be 1.0? Or am I misunderstanding your post?

[EmiliaEQ]

You agree with me that haste/slow dont affect PPM
You agree with me that extra swings (via skills/aa) dont affect PPM

Thus Extra Swings = Haste/Slow = No affect on PPM
Which means there is an innate coefficient that is here to lower
down the proc rate per swing, do compensate with the increase of swings du to the haste.

Why did i use something as ACC ?
Because there need to be a variable class dependant to lower
the proc rate depending on the class using it.

Primal 2HB = 20 delay & 2 PPM
A ranger will do 30 swings in 1 minute, and get 60 attacks (assumeing AAC of 2)
A cleric will do 30 swings in 1 minute, and get 30 attacks.
(they get no dual/tripple atk)

Yet both of you will have a 2PPM proc rate.
To get the same PPM the cleric will have a double proc rate as yours.
Or actually you would have to have half his proc rate.

My idea is that each weapon has a % chance to proc per attack
(NOTICE i said attack not swing).
There has to be some "hidden" coefficient that will lower the
proc rate to compensate for the increased swing rate.

So in theory with ZERO HASTE:

Primal 1HB proc rate per swing = 6.6% / AAC

Since clerics AAC is 1 , their proc rate remains 6.6%
Since rangers have multiple attacks their proc rate is much lower.

The overall proc rate in PPM will be the same.

How do we determine the AAC ?
Well its the average number of attacks per swing made by the character.
Its based on Dual attack skill + some AA.
And is easily parsed.

[Ladas Sindarin]

I wasn't questioning your reasoning, I was questioning your examples which didn't follow your reasoning.... which you have so conviently edited to fix without noting.... while you are editing your posts, you might go through and make sure you use the terms "Swings" and "Attacks' consistantly...

[EmiliaEQ]

Yes its a pain not to mix them up :)

[Feysboil]

I have to believe that, in the interest of simplicity in coding, procs work off a simple percentage chance to fire per attack round, with modifiers for things like DEX and AA abilities. I think coding anything related to combat in this game to be based on occuring on a "per minute" basis is absurd. Of course, that may only mean that I think the developers are absurd, in the event that that's how they've really coded it.

The other alternative is that I've misinterpreted the meaning presented here for the conclusion that there are "2 procs per minute."

In essence, I think having hard-coded 2 procs per minute doesn't make sense because it's predictable, but not in a manner that's consistent with any style of event-driven coding I'm familiar with.

Basically, what you have is a clock on the server, and that clock drives everything to do with straight melee. Based on weapon delays and haste/slow, you get an attack attempt at concrete, fixed intervals. Depending on several factors, you might get 10 attacks in a minute, or you might get 30, or anywhere in between. I'd rather believe that secondary effects like procs occur not based on a separate timer, but as a result of each attack event along the main timeline.

Imagine combat as a line...

- - - - - - - - - - - - - - - - - - - - - - - - - - - -

Each "-" is a pulse of the combat clock (some block of seconds, for simplicity's sake, we'll say 1 second per pulse). Let's say your weapons give you 1 attack attempt every 3 pulses. The timeline looks like this (let's just assume for now we're talking about main-hand only).

- - * - - * - - * - - * - - * - - * - - * - - * -

Where "*" is an attack.

Now, on each of those attacks, several things may happen:

1) the attack may miss

2) the attack may hit for normal damage

3) the swing may result in a critical/crippling blow

4) double/triple attack may trigger

5) weapon may proc, if applicable.

Typically, these events are not mutually exclusive. For instance, you may miss, but succeed on your double attack roll, resulting in a second attack roll, which may hit, and which may also coincidentally land as a crit, and trigger a weapon proc. Each of those secondary effects (double attack, critical/crippling blow, proc) are additional rolls against individual percentages for each.

I think it's just a coincidence and an effect of averaging over many trials that the "rate" appears to be two procs per minute. I have seen long periods time go by without seeing a proc go off, and I've also seen that same weapon proc half a dozen times or more in a minute.

So you might see something like this on that line:

M=miss
X=hit
2=double attack
C=crit/cripple
P=proc (assuming includes a hit)

- - M - - X - - X - - 2(XM) - - P - - 2(PX) - - M

That would indicate the progression:

Miss
Hit
Hit
Double attack with a hit and miss
Proc
Double attack with a proc and a hit

[Cily]

Feysboil -

Er... no

There is really no argument about the 2 proc per minutes anymore.

I have seen 10 hour parses of someone using a slow 2 hander, unhasted and it produced 2 procs per minute.

I have my own parses of many hundreds of hours using a hasted, 18 delay weapon, and it produces 2 procs per minute.

To assume this is just coincidence, and that I got unlucky during that several hundred hour stretch is ludicrous.

Also, the effect slow has on Earthshaker is plenty enough evidence for me.

Cily

[Feysboil]

You're still talking about averages over many many repetitions. I would rather believe that the proc rate is inversely proportional to weapon delay, such that the average proc rate is twice per minute.

I'm not disputing that many tests show an average of 2 procs per minute. I'm suggesting that believing it's as simple as them saying "Weapons with procs will proc twice a minute - programmer person, make it so." The proc rate isn't tied to time, it's tied to the inverse of delay. That's why faster weapons proc less, and slow weapons proc more, on a per-swing basis.

[Gimmel]

You're still talking about averages over many many repetitions. I would rather believe that the proc rate is inversely proportional to weapon delay, such that the average proc rate is twice per minute.

Yup, that's what I've been saying since the cleric hammers first came out. But people seem to have difficulty understanding that concept, but grasp 2 procs/minute more readily. Same result.

[Grumblethunk]

Any proc rate discussion that doesnt acknowledge the huge proc tests the Shaman did (and still have up at the cruicible) is like a heart study that ignores Framingham.

ie, its re-inventing the wheel.