WHAT IS A PROC RATIO ? :
Every time you use a skill in order to proc an effect (fear on it or bleed for example) you have to multiply the probability given in the description of the effect by the proc ratio of the skill you use to have the real probability that the effect occurs using that skill.
WHO WAS THE FIRST TO MEASURE THE RATIO :
That man/woman : http://us.battle.net/d3/en/forum/topic/5149154265
WHERE MAY I FIND OTHER CLASSES 1.07 RATIOS :
Other classes from ZzEzZ : http://us.battle.net/d3/en/forum/topic/7004036888?page=1
WHERE MAY I FIND A SUMMARY OF ALL INFO ABOUT WIZARD :
The wizard build compendium from AS00 : http://us.battle.net/d3/en/forum/topic/7004037930
dot : damage over time
tick : one damage of a damage over time
cast speed coef : you have to multiply your attack speed by this value to know the number of time you can cast this spell per second.
tick speed coef : When the skill got this parameter that mean you need to use the generalized VOLTARA formula to calculate the number of tick. (please be aware that the character sheet round some data so you may want to recalculate your own data to get an accurate result.
no anim : the spell could or could not trigger an animation but does not need one.
channel : mean that you can maintain the incantation of the spell
the text in italic concern point that need more testing.
Enough cheap chat here are the ratios :
Rune 1 - Obliteration : 0.333 [1/3]
Rune 2 - Arcane Orbit : 0.250 [1/4]
Rune 3 - Arcane Nova : 0.167 [1/6]
Rune 4 - Tap the Source : 0.333 [1/3]
Rune 5 - Celestial Orb : 0.200 [1/5]
Rune 1 - Disruption : 0.200 [1/5] per missile (speed coef : 3)
Rune 2 - Death Blossom : 0.600 [3/5] per missile (speed coef : 3)
Rune 3 - Arcane Mines : 0.500 [1/2] per mine (speed coef : 1)
Rune 4 - Power Stone : 0.200 [1/5] per missile (speed coef : 3)
Rune 5 - Cascade : 0.160 [4/25] per missile (speed coef : 3) - additionnal missile : 0.160 [4/25] per missile
Rune 1 - Grasping Chill : 0.0125 [1/80] (24 time over 6 seconds stack)
Rune 2 - Frozen Solid : 0.0125 [1/80] (24 time over 6 seconds stack)
Rune 3 - Snowbound : 0.0125 [1/80] (24 time over 6 seconds stack)
Rune 4 - Stark Winter : 0.0125 [1/80] (24 time over 6 seconds stack)
Rune 5 - Unrelenting Storm : 0.0125 [1/80] (32 time over 8 seconds stack)
Rune 1 - Convergence : 0.111 [1/9] per tick
Rune 2 - Chaos Nexus : 0.111 [1/9] per touch tick - additionnal beam : 0.111 [1/9] per tick
Rune 3 - Volatility : 0.125 [1/8] per tick - explosion : 0
Rune 4 - Entropy : 0.167 [1/6] per tick
Rune 5 - Intensify : 0.167 [1/6] per tick - intensified : 0.167 [1/6] per tick
Rune 1 - Chain Lightning : 0.083 [1/12]
Rune 2 - Forked Lightning : 0.167 [1/6] - charged bolts : 0.167 [1/6] per charged bolt
Rune 3 - Lightning Blast : 0.167 [1/6]
Rune 4 - Surge of Power : 0.250 [1/4]
Rune 5 - Arc Lightning : 0.100 [1/10]
Rune 1 - Mistral Breeze : 0.125 [1/8] per tick (stack)
Rune 2 - Gale Force : 0.125 [1/8] per tick (stack)
Rune 3 - Raging Storm : 0.125 [1/8] per tick (stack) - big twister : 0.125 [1/8] per tick (stack)
Rune 4 - Wicked Wind : 0.125 [1/8] per tick (stack) *
Rune 5 - Storm Chaser : 0.125 [1/8] per tick (stack) - purple tornado non dependent of the number of stack : 0.125 [1/8] (no anim)
* : APO and SHANDLAR testing : http://i.imgur.com/t2vsp.png
VOLTARA gets the formula here : http://us.battle.net/d3/en/forum/topic/6794871641?page=9#168
Rune 1 - Unleashed : 0.333 [1/3]
Rune 2 - Time Bomb : 0.333 [1/3]
Rune 3 - Short Fuse : 0.333 [1/3]
Rune 4 - Obliterate : 0.125 [1/8]
Rune 5 - Chain Reaction : 0.111 [1/9] (3 time)
Rune 1 - Shatter : 0.167 [1/6] - nova on kill : 0.000 
Rune 2 - Cold Snap : 0.167 [1/6]
Rune 3 - Frozen Mist : 0.050 [1/20] - dot : 0.050 [1/20] (27 time over 8 seconds stack)
Rune 4 - Deep Freeze : 0.167 [1/6]
Rune 1 - Bone Chill : 0.167 [1/6]
Rune 1 - Charged Blast : 1.000 
Rune 2 - Split : 0.350 [7/20] per missile
Rune 3 - Penetrating Blast : 0.333 [1/3]
Rune 4 - Attunement : 1.000 
Rune 5 - Seeker : 1.000 
Rune 1 - Molten Impact : 0.125 [1/8] - dot* : 0.125 [1/8] (3 time over 3 seconds stack)
Rune 2 - Star Pact : 0.10625 [17/160] - dot* : 0.10625 [17/160] (3 time over 3 seconds stack)
Rune 3 - Meteor Shower : 0.050 [1/20] per small meteor (7 meteors) - dot* : 0.050 [1/20] (2 time over 2 seconds stack)
Rune 4 - Comet : 0.125 [1/8] - dot* : 0.125 [1/8] (3 time over 3 seconds stack)
Rune 5 - Liquefy : 0.10625 [17/160] - dot* : 0.10625 [17/160] (3/8 time over 3/8 seconds stack)
* : The first tick of the dot appears right at the impact.
Rune 1 - Numb : 0.333 [1/3] per tick
Rune 2 - Snow Blast : 0.333 [1/3] per tick
Rune 3 - Cold Blood : 0.333 [1/3] per tick
Rune 4 - Sleet Storm : 0.1875 [3/16] per tick
Rune 5 - Black Ice : 0.333 [1/3] per tick - ice : 0.333 [1/3] (12 time over 3 seconds, stack)
Rune 1 - Explosive Bolts : 0.800 [8/10] per bolt - explosion : 0.800 [8/10]
Rune 2 - Fire Bolts : 0.8 [8/10] per bolt
Rune 3 - Piercing Orb : 0.500 [1/2]
Rune 4 - Lightning Affinity : 0.8 [8/10] per bolt
Rune 5 - Living Lightning : 0.200 [1/5] per charged bold
Rune 1 - Deep Cuts : 0.140 [7/50] (3 time) - dot : 0.140 [7/50] (1 time for LoH, 30 time over 3 second for CM, unkown for everything else, probably 0), all or none tick crit
Rune 2 - Impactful Blade : 0.140 [7/50] (3 time)
Rune 3 - Siphoning Blade : 0.140 [7/50] (3 time)
Rune 4 - Healing Blades : 0.140 [7/50] (3 time)
Rune 5 - Thrown Blade : 0.080 [2/25] (3 time)
Rune 1 - Impactful Wave : 0.167 [1/6]
Rune 2 - Force Affinity : 0.167 [1/6]
Rune 3 - Forceful Wave : 0.167 [1/6]
Rune 4 - Teleporting Wave : 0.167 [1/6]
Rune 1 - Exploding Wave : 0.125 [1/8] - small wave : 0.125 [1/8]
Diamond Skin / Diamond Shard : 0.167 [1/6] (no anim)
Frost Armor / Frozen Storm : 0.125 [1/8] (6 time over 3 seconds, no anim)
Mirror Image / Mocking Device : 0.500 [1/2] per decoy (2 decoys)
Teleport / Calamity : 0.167 [1/6]
Archon / Melee Attack : 0.500 [1/2]
Archon / Ray Attack : 0.100 [1/10] per tick (channel tick speed coef : 3)
You may experience some effect trigger to happen with some skills regardless of their proc scalar. Probably because of pseudo RNG. A good example of pseudo RNG is the targeting system of death blossom.
CC.R per monster hit chance to trigger CM out of using that skill
or if you prefer :
R per monster hit chance to trigger CM out of scoring a critical with that skill
Concerning dot with very short time between each damage : it appears that the game keep the random value used to pass the ratio check for each damage dealt by the dot. That mean that each damage has only to pass a critical check. This lead to an absurd amount of CM trigger on dot if you are lucky enough to pass the proc check.
R per monster hit chance to trigger a flash of insight
LoH.R per monster hit
R.AC.CC arcane power per monster hit per casting of the spell.
Or if you prefer
R.AC arcane power per monster hit out of a critical from that spell
[EDIT 19/04/2013 : Update on deep cut, the answer is 42!]
Edited by Guybrush#2456 on 19/04/2013 13:51 BST
Why don't you post this in that original thread?
Because this was a us thread and we are european ! :D Only kidding...
I made a new thread so you don't have to search all the pages to get the ratios.
Here's an interesting one:
You are never sure when you are talking about proc ratio. Maybe I made a mistake copying the value. I will test it again.
Also a mistake could have been introduce because for some of them i just made a "per touch" test. As I said at the beginning of the post : if you made some conflicting testing or you just have doubnt one some value : post it there so we/I/somebody can test it more and conclude.
[EDIT : yes the 0.500 ratio on piercing orb is confirmed]
Edited by Guybrush#2456 on 24/08/2012 21:39 BST
Very interesting. For disintegrate - Volatility, are you absolutely sure the explosion triggers? I absolutely think it should given the lower proc rate than a runeless of Intensified version seeing as the actual beam is no larger than theirs, but no one previously found that it did, so I just want to clarify. With the chance of the small area explosion only triggering on 35% of DEATHS rather than hits, it makes sense that the proc rate for the beam would only be reduced if the area effect can also proc.
Typos/spelling errors abound in your unofficial list, you might want to tidy that up seeing as you may just have become the host of one of the newest pages for people to be glad they still have in their cache (so that you can just type proc ratio in the URL bar and have it pop up as the first in the list). Well done on that! (The updated details, I'm not teasing for the typos!)
It'd be better to change the results from being rounded off because it's less accurate and you basically know what they really are anyway. Most of them are obvious fractions:
1, 1/2, 1/3, 2/3, 1/4, 1/5, 1/6, 1/8, 1/9, 2/9, 1/10, 1/12, 1/20, 1/50 and 1/100 and that's about all of them covered.
So any "0.33" and "0.35" (Split, you say) are pretty much guaranteed to be 0.333 (1/3), "0.670" is 0.666 (2/3), 0.165 is really 0.166 (1/6), 0.110 is 0.111 (1/9), 0.220 is 0.222 (2/9), 0.085 is 0.0866 (1/12) and I think the others are all even more obvious or all already perfect fractions in decimal form. Most people can recognise most of these fractions when in decimal form, maybe just 0.111 or 0.083 could throw a few people here or there, but to avoid any confusion about whether they are or not, you could check that Split is in fact 0.333 rather than 0.35 and feel confident that all of the others are common fractions. Three decimal places is best for clarity in my opinion, or you could go for two as a base and then always add an extra 1/3/6 in the case of those with recurring numbers like 1/3, 1/6, 1/9.
(Runeless) : 0.250 - This was previously the case, but it hits at twice the normal rate of fire, so compared to most rates, this is like 0.5, meaning they expect it to hit 2 on average per tick, which is nice as it is harsh when they seem to base things on the maximum something is likely to hit. Dekkar found 0.166, but I think someone later in the thread found 0.25 as you have, so maybe it was changed early on from expecting three per tick.
Rune 1 - Chain Lightning : 0.085 - Would be 0.083(333-->) as that is 1/12. They expect it to hit one in sixth on average over the normal attack period (given the double attack rate). This is an example of them being very (too) harsh as it can only hit six at maximum, while some area spells that might be predicted to hit 6 are at least capable of hitting more! The original thread shows that Dekkar found 0.166, which would be only three per tick, maybe fine, but maybe a little generous, so they seem to have later nerfed it by halving it, which seems like overkill seeing as they could have easily chosen 0.125, expecting double that of Electrocute, for four per tick and eight per period. If you're reading, Blizzard, shouldn't this be 0.166? *Bats eyelids* Pleeeeease?
Rune 2 - Forked Lightning : 0.165 - charged bolts : 0.165 per charged bolt - Makes sense as, like Volatility, it has the same basic attack but then potential but unreliable area effects on death that can then also trigger proc.
Rune 3 - Lightning Blast : 0.165 - unchanged, expected to hit three per tick, although the new graphics (which sadly seem more realistically sized and shaped now) make it look like less will be hit. Hrmm. I'll miss my giant zapping arrowheads.
Rune 4 - Surge of Power : 0.100 - was 0.166 apparently, it seems a little odd that it now has less than the runeless and Forked versions to me.
Rune 5 - Arc Lightning : 0.100 - Huh? They expect this to regularly hit 5 per tick? Doesn't this have an area similar to normal Spectral Blades, only narrower close to the body and wider at the end of the cone? 0.200 over a normal attack period (1 second at 1.00 attack speed) compared to 0.66 seems fishy. Is there something we don't know about this spell?
It's a little odd to me that they really do still let Deep Cuts proc on the Bleed effect triggering (immediately and once, rather than on each tick) and I think it's a shame that Ray of Frost really is only 0.333 per tick rather than 0.500, because it has zero area of effect and sometimes doesn't seem to do damage against a target you are focussing on. Whereas Disintegrate can hit a target you are aiming the beam at by clicking IN LINE with a target rather than directly on it, Ray of Frost usually seems to behave differently by sometimes missing when I do that but always hits normally when I have the cursor directly over the unit. I wish they'd either fix this or apply an extremely small (0.025 range or something) area of effect to the end of it. Until that happens, at least, it seems pretty silly that RoF procs less over a normal time period when it is only single target.
Anyway, I've been looking at the other proc page for hours and just suddenly found this, to my great delight; it should be a great updated thread. The only thing is that your post may be copy+pasted into the other thread and people might just continue discussing it there instead, but I don't know.
Edited by Foxpocalypse#6905 on 24/08/2012 23:16 BST
Yes... I am french and you all know how messy is our english :p. I'll try to correct as many mistakes as I can each time I update the post.
I am not totally sure about that but I will remake the test with a different life on hit value so we can be sure.
I was sure, because i made some test in act 1 normal where i slowly approached the beam to a pack of monster to be sure to hit only one and keep the data if he explode. But now that i re do it : it does not proc life on hit. Maybe I was pretty sleepy. Sorry about giving you false hope anyway.
I re-tested it... now i get 0.250 I probably did a nasty copy/paste.
I re-tested it : it still give me a proc ratio of 1/10. Moreover the spell is not a dot and it does not hit several time like deep cut. The aoe seems bigger that spectral blade though.
Thank you for the feedback Foxpocalypse !
[EDIT : oops i did not put the right quote about volatility : fixed.]
Edited by Guybrush#2456 on 25/08/2012 09:02 BST
No worries! Okay, so on Arc Lightning, this is just odd from Blizzard as far as I can tell, unless I'm missing something about the ability that makes it good enough so that it shouldn't proc often (or it hits more on average than I think).
You misquoted by quoting the Surge of Power paragraph twice, but assuming the first was on Volatility, DAMN!!!!!!!!!!!!!!
It should proc on the explosion or else have the same proc per tick for the beam as without a rune or with Entropy and Intensify. Change it Blizzard! :"(
On Surge of Power, yeah, it would make the most sense for it to have the same as the normal one, though some people debate that runes that make the spell cheaper (which could also iclude runes that give back AP) should proc less because you can cast them more often. The fact is though, that's clearly not how Blizzard sees it!
As for the fractions, you can see on the original list that pretty well all of them fall into the decimal values of common fractions, but it's good to be sure, so I won't stop you testing again if you wish to do so!
How much Life on Hit did you have? For testing, people previously discussed using 1000 if possible, but that number doesn't evenly divide by 3, which *might* make things less clear. With 1200, things with a 33% proc chance would give 400, one ninth (some Convergence☆Disintegrate ticks) would give, oh...133.33, hmm, well maybe 900 is better then. :P
The idea was that you would mostly get clear ###.00 or ###.50 (maybe ###.25) types of results, but really, whenever you get (seemingly) recurring numbers from a large starting number it's extremely likely to actually be recurring (like those fractions).
Anyway, attack speed can affect the numbers displayed by the channelled spells, but testing with a weapon of base speed 1.00 with no AS modifiers sounds like the best starting point to me. What did you use?
I'd also like to add that those 900+ life on hit amounts for testing would be easier these days, for two handers at least. :P I just got a 1800+ Life on Hit weapon upgrade for my wizard. :)
I'd also like to link a particular page of the original thread that I didn't notice getting as much attention but personally found especially useful:
Nyphur's first post from June 1 (2012, in case we look back to this post late next year ^_^)
From my experiences, I'm sure his second theory on tick rates is the right one.
Edited by Foxpocalypse#6905 on 25/08/2012 03:28 BST
Ran trough a spell check and corrected some grammatical errors. I've also taken the liberty to rewrite small parts of two or three sentences to make them a bit clearer. It's a bit early and I've just had my coffee so I hope I haven't missed anything. ;)
Hopefully your awesome work will be shared across theory crafting sites/pages so it never gets lost. :) Very well done!
As every theory-crafter knows there was that incredible thread about ratios back in 1.02 :
Hydra does not proc any life on hit, or critical for that matter. It used to proc life on kill when the hydra was dispelled but it does not anymore.
Edited by Guybrush#2456 on 26/08/2012 10:22 BST
Hey guys... I am having some trouble testing the probabilty that critical mass have to trigger. I Tested it with spectral blade / deep cut and it appears that the critical mass trigger is delayed compared to the casting of spectral blade... But when I use magic missile CM trigger right away if a critical is scored.
Life on hit wise : the global proc ratio is 0.888 per cast there is no doubt about it
I am also sure that the ratio of the first blade is 0.222 because it has been tested on monster that die in one hit.
CM wise however it have a 0.222 ratio 3 time like every spectral blades but i am not sure of the 4th proc.
The dot duration is 3 seconds so if it procs we should get several green number over our head : just like meteor. But there is only one green number and it is accurate. Arcane dynamo wise the flash of insight trigger also per blades and the max i could have so far was 3 flash out of 1 cast. HOWEVER and that is where is start to become even more interesting : Critical mass proc during the dot !
I don't understand it because I thought that life on hit, critical mass, arcane dynamo, every effect that you get on your gear share the same proc ratio. Apparently there is some inaccuracy on that assumption.
The random value that the system roll appears to be kept for a said period of time (probably for optimization) so if the first tick of the dot proc : a lot of proc is incoming ! Just like blizzard by the way. It is not troubling for stun on hit because it will just refresh the fear for minus thatn 1 second. But it is troubling for critical mass !
According to my experience almost all random function you can have on programs is based on the "clock" of the processing unit. Basically to be executed by a processing unit a program has to be translated into very basic sentence : put that value in the stack, get the first value of the stack, get the value stored here, ... and so on. Each of this operations take a known amount of time from the beginning of the task to the end of the task. That is where the word "overclocking" come from by the way. If it appears unclear to you let's just say that computer create "random" number out of totally non-random number in a logical way. I don't want to exploit the game so I will probably stop wondering about how the game get the random value.
[EDIT : clarify my point, and also correcting an horrible mistake i made]
Edited by Guybrush#2456 on 25/08/2012 20:36 BST
Threats of violence. We take these seriously and will alert the proper authorities.
Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.
Harassing or discriminatory language. This will not be tolerated.