Talk:Spell power coefficient

Does Devouring Plague actually get 100% instead of 80% ((24/15)/2). --bfx 15:24, 23 March 2007 (EDT)
 * Im almost certain I remember a blue post saying coeffecients consider unmodified spellcast time. So that faster fireball/frostbolt wouldnt lower the coeffecient. —The preceding unsigned comment was added by.
 * In the case of those spells, you should look at the talents again.--Sky (t · c · w) 02:14, 5 April 2007 (EDT)
 * Yea just noticed, i just remember a blue being adament about the situation long ago, but go go nerfstick i suppose. Thain 02:16, 5 April 2007 (EDT)
 * Indeed, go go nerfstick. /sigh. ---> is a Mage.--Sky (t · c · w) 02:20, 5 April 2007 (EDT)
 * Fireball/frostball changed in patch 2.3, before it got -10% from talents but now id has still 100% when you have 5/5 is speed increase talents, so they should be changed here accordingly to it.--Vidinufi 14:21, 18 November 2007 (UTC)

What's with all the non-exact figures?
The mage ones, for example. Some spells are clearly just 1.5/3.5 but there's 42.86% and 42.73% in different spells. It's 42.86% if there's nothing weird about the spell.

It looks like it's the result of repeated testing and statistical averaging. No need for that on spells which are "obvious". --Mekkapiano 12:35, 1 July 2007 (UTC)

They aren't always predictable as 1.5/3.5, sometimes the observed values are close to the formulaic versions, but don't match them close enough to be exactly predicted, great examples are AoE coefficients that can't quite be justified by formulas. Zaldinar 04:35, 14 November 2007 (UTC)

Regrowth Coefficient


The regrowth coefficient was listed at approximately 28-29%. Regrowth is a 2 second cast (normally 4/7) followed by a 21 second HoT (normally 1.4). The HoT gets exactly half of what it should, 0.7 of your +heal. Dividing 4/7 by 2 gives 28.57%, right in the experimentally determined range. I think it is reasonable to conclude that this is how they arrived at the coefficient for regrowth, the spell is in two parts and each part gets half the amount of coefficient that a spell of that type normally would. It seems reasonable, then, to surmise that 28.57%, or two sevenths is the precise coefficient for the initial burst of regrowth. --L337n00b 18:27, 7 July 2007 (UTC)

Indeed - I'm often annoyed at the preference to supply a statistical figure rather than just the most obvious explanation. It should be the reverse process: apply reasoning to get a figure which looks right given trivial testing, and THEN try to disprove it by finding any case which disagrees. Sometimes a 3 second spell really is just 3/3.5... --Mekkapiano 10:21, 10 July 2007 (UTC)

The split is not 50/50. 4/7 and 1.4 are correct, but the splits are 4/7 / (4/7+1.4) = 16.6 and 1.4 / (4/7+1.4) = 99.4 --Erdluf 12:59, 6 September 2007 (UTC)


 * 4/7 / (4/7+1.4) = 29.0% and 1.4 / (4/7+1.4) = 71.0%... --ElginRoko 15:35, 13 February 2008 (UTC)

Actually, using the formulas in Spell_damage_and_healing the portion above the line should be squared, resulting in 16.6 and 99.4. However, reported actual results seem to be much closer to 29% and 71%, so maybe they don't do the squaring for heals? If you don't do the squaring you get very close to the result for Lifebloom hot, although the value for the bloom result is too large. You do need the squaring to get the right values for Moonfire.

Anyway, I'm inclined to believe the correct numbers for Regrowth are very close to 29% and 71%. I will do some testing, and update, if nobody beats me to it. --Erdluf 18:27, 19 March 2008 (UTC)

My testing showed the Regrowth HoT to be very close to 70% (between 60.9% and 70.1%). The direct heal is about 30%. My average was slightly higher than that, but not enough samples to be sure. I updated the article for this. --Erdluf 12:18, 20 March 2008 (UTC)

Priest coefficients
I did some tests with Holy Fire. Without any point in Holy talents, my results were that the DD receives between 83% and 87% and the DoT receives approximately 16.5% (there was a slight fluctuation from one HF to another). These values lead me to believe that the actual value for the DoT is somewhere around 16.5%, whereas the DD receives either 85.71% (3/3.5) or ~83.5% (1 - ~16.5%). Does anyone else have some values? So far, this complies neither with the previous assumption of 75%/25%, nor with the actual proportions. --bfx 10:39, 27 July 2007 (UTC)


 * I tested and came up with the same values. Editing them into the page now. --KoenPater 12:21, 9 October 2007 (UTC)

I've just noticed that Touch of Weakness is listed with a coefficient of 10%. Currently, I have no Undead priest, but I believe it was and still is 43%. --bfx 05:37, 30 August 2007 (UTC)
 * Nope, just tested it and 10% seems right. --KoenPater 12:21, 9 October 2007 (UTC)

Devouring Plague
Testing now, on the 3.0 PTR, and appears now that Devouring Plague is been changed from 80% coefficient, to a 24 seconds duration coeffcient; so ~160%. I'm no expert at math but if someone could double check it for me, it seems to be correct.

Fr0
 * My tests show that the coefficient for Devouring Plague hasn't changed. Its damage is increased by 80% of the spell power value, as is its healing. You might have forgotten to refrain from using any talents while testing--which, of course, would lead to a significantly higher coefficient. --bfx (talk) 12:37, 23 September 2008 (UTC)
 * With the release of Patch 3.0.2 I have added the untested "flag" because it seems to have received a new coefficient. --bfx (talk) 05:49, 15 October 2008 (UTC)

Holy Nova
It appears that no one has tested the coefficient for Holy Nova. Samalander 23:45, 16 December 2007 (UTC)

I tested it on the PTR 3.0 as well, and it seems to be closer to ~35%. Please double check for me. Fr0

Mind Flay
Note: My own preliminary testing delivered a coeficient of 67,5% for Mind Flay. Not to debase the work already done here, but i'm not so sure how reliable this values are. --WurmD
 * Did you test it without any Shadow talents? This could easily lead to wrong results. Anyway, it is still possible that these values are not correct. Specifically, Mind Flay's coefficient once was mentioned by a CM; I haven't heard of any changes to it, but that doesn't mean it didn't happen. --bfx 05:33, 30 August 2007 (UTC)

Talented coefficients
Shaman, Warlock, and Druid coefficients are listed with untalented values first, and then talented values in parenthesis. I've updated the Druid coefficients (damage spells only) to include all applicable talents (Improved Moonfire, Moonfury, and Wrath of Cenarius). At some point someone had undone some of the Moonfury contribution, for reasons unknown --Erdluf 13:04, 6 September 2007 (UTC)
 * It was me who did this. The reason is that the other classes only have talents included which actually change the benefit of spell damage, not the total damage (as talents like Moonfury do not specifically affect spell damage values). The only spells which have comparable talents included are Frostbolt and Fireball, probably due to the unusual mechanic of their respective talents. The only talent of the Druid which actually belongs here (at least as it is now) is Wrath of Cenarius. However, I will not change it again until someone else tells their opinion about it. I personally don't like them here, as for specific classes (especially the three main spellcaster classes Mage, Priest and Warlock), it would horribly mess up the page (due to every single spell being affected by talents; furthermore, for example a Priest has different options, such as using the damage increasing talents from the Shadow or the Discipline tree). --bfx 15:18, 6 September 2007 (UTC)
 * I'd like some method of getting at "final" spell coefficients. For instance I believe that prior to 1.8, Moonfury did not scale, and you still see websites with the old forumulas.  I was also under the impression that some healing related talents still don't scale, but that may just be a faulty memory at work.  For the most part I guess I'm really more interested in seeing a consistent, and up-to-date (ie level 70 values) version of Spell_Damage_Comparison. Its just that page is such a mess (what is the definition of DPS) that it would be hard to tackle.  --Erdluf 17:04, 6 September 2007 (UTC)

Immolate is listed as receiving 27.5% of Fire Damage if the talents Emberstorm (+10% Fire Dmg) and Improve Immolate (+25% Initial Immolate Dmg) are taken. I'm assuming 20 x 1.1 x 1.25 was used to reach this conclusion. It is actually incorrect. The calculation for total damage of the initial Immolate spell is ((Base + Co-eff of 20%) x 1.1 x 1.25). I mention this as I have tested it with my Warlock. 819 Fire Damage, 5/5 Emberstorm and 5/5 Imp Immolate. Using the 27.5 co-eff and using the standard ((Base + Spell Dmg) x 1.1 x 1.25) formula, I should be hitting for 759. I am actually hitting for 675, which is using the co-eff of 20%. Co-efficients should only be changed where a talent increases Spell Damage used by the spell (eg Shadow and Flame, Emp Corruption) and should not be changed where a talent increases the total damage of a spell (eg Emberstorm, Shadow Mastery).````--Razal 13:50, 3 November 2007 (UTC)

Consistent look
This page finally needs a consistent look.
 * Should DoTs/HoTs have their total value, their tick value or both listed? The same goes for spells with charges (e.g. Lightning Shield).
 * Should talents which increase overall damage (and not the benefit from spell damage) be included?
 * Should the highest ranks be used, or only a value independet of rank (for example, Corruption has only the total damage coefficient for the highest rank, whereas Entangling Roots has the value per tick)? --bfx 15:44, 6 September 2007 (UTC)
 * I certainly agree that a consistent look would be good. I'm usually more interested in the total spell coef.  However roots may be a special case (at least among the Druid spells).  The coef/tick is constant at all spell levels (at least above 20, and not downranked), and in PvE/PvP.  The number of ticks varies by spell level.  It also changes between PvE and PvP.  Top rank is nine ticks for PvE, and currently four ticks for PvP.  In 2.2 it will drop to 3 ticks for PvP.  I guess Force of Nature is similar, in that nobody is sure exactly how many hits are possible from a single cast, and you'll probably never get the upper limit in practice.  --Erdluf 16:35, 6 September 2007 (UTC)


 * DoTs and HoTs should have their coefficients listed as part of their total value unless the coefficient is calculated on a per tick basis (I don't know of any that are). This is particularly crucial given that different spells have different ticks (ie. Lifebloom ticks every second, Moonfire every 3 seconds).  Additionally, splitting the total coefficient up and distributing it over each tick lends itself to far greater inaccuracy.  Given the significance and very fluid nature of this page's information, a couple of theorycrafters really should get a project going that focuses exclusively on maintaining it.  They could establish standards and practices and testing methodology (I don't think the look needs to really be messed with, as it's already very well organized).K10wn (talk) 17:24, 4 October 2008 (UTC)

What about hunters?
I have never played one by myself and although I have never seen oriented-hunter/hunters dedicated gears including spell damage, their damages are both from physical and arcane schools; now they share mail with shamans so mail can include spell damage, not to mention the possible use of leather, clothes (likely to have spell damage), and spell power enchantments. Is it by lack of information from the community? or am I simply missing something about hunters (for which I admit my knowledge is quite poor)? In both cases, I'd tend to think wowwiki, as a exhaustive wiki, should include knowledge from an aware hunter, or possibly (if this is the case) just a single sentence to state that hunter arcane spell damage do not receive any increase of damage from gear. --Staarkali 01:41, 1 January 2008 (UTC)


 * I'm not sure if Arcane Shot will benefit from any +Spell Damage but I do know that Mend Pet gains (or used to gain) from +Healing. I once saw a video on YouTube of a hunter who stacked +Healing and his Mend Pet ability was ticking at over 2000. The video showed portions of the hunters solo fight against the old world outdoor raid Dragons. --Darxide 01:50, 1 January 2008 (UTC)

Paladin Retribution Aura?
Does a Paladin's Retribution Aura gain from any +spell damage? Or possibly from Judgment of the Crusader? I have noticed that fully raid buffed, my Ret Aura will gain +1 and sometimes +2 to the damage it reflects. Possibly Ret Aura has an extremely small coefficient? --Darxide 01:52, 1 January 2008 (UTC)


 * In my experience, Retribution Aura is affected by buffs/debuffs that increase spell damage by a percentage (zone buffs, 1-H weapon spec, etc.), but not by spell damage. --User:Hypehuman 29 April 2008


 * In wrath it does scale with spellpower giving approximately 6.5% of it as extra damage. Muxecoid (talk) 16:28, 15 November 2008 (UTC)

Lifebloom Coefficients
The talented coefficients for lifebloom are all wrong. Empowered rejuvenation adds 10.9% to the HoT not 20%. Pretty easy to test for any resto druid, the 20% addition is way off. The reasoning is nicely revealed by the 2.4 PTR patch notes - "Empowered Rejuvenation: This talent now properly affects the final heal from Lifebloom." The other 9.1% should be being added to the final heal, but isn't being. --ElginRoko 15:28, 13 February 2008 (UTC)

I had put those numbers in. I believed that Empowered Rejuvenation would add 20% to the coefficient (similar to the way Wrath of Cenarious adds 10% to the Wrath coefficient). Testing clearly shows that ER multiplies the coefficient by 1.20. --Erdluf 18:30, 19 March 2008 (UTC)

The tooltip of Empowered Touch/Wrath of Cenarius says that it "... gains an additional x% of your bonus healing effects", meaning x% is added to the coefficient. If the tooltip says "The bonus healing effects of your spell is increased by x%", the coefficient is multiplied by x%. I had only had thoughts of this before, but it is confirmed now. --Treekebab (talk) 09:48, May 27, 2010 (UTC)

Holy Nova Coefficients Added
I just did some tests on Holy Nova and edited in my results. I found that the coefficient is 16% (±0.3%). I've heard that the coefficient for the healing was 16%, and my results are consistent with this, although they had a much larger margin of error. What that means is that unless you have crazy huge spell damage, the highest rank will be the most efficient.

To get the 16%, I spammed holy nova on some mobs and recorded the minimum and maximum damage until the difference between the two was the same as the difference between the minimum and maximum listed in the tool tip. I took the difference between the recorded damage and the tooltip damage (assumed an error of ±1) and divided it by my spell damage. For the healing portion, I did not test long enough to get the actual minimum and maximum (the difference between them was quite small), but the min and max were within the range of values that come from assuming a 16% coefficient for the healing. --User:Hypehuman 29 April 2008

Seal of Vengeance Coefficients Added
I just did some tests on Seal of Vengeance. I was able to confirm the DoT and Judgement coefficients that were already there, and I added in the coefficient on the 20 bonus damage that is incurred when the DoT is refreshed and there are already the maximum number of applications. All the damage numbers alternated between two adjacent integers, and so I assumed that the "real" value was somewhere between the two. That's why the numbers are all something.5, and I took the limits of the value as ±.5 from there. Here's my raw data:

--User:Hypehuman 29 April 2008

This Page Needs Revisiting
Every coefficient on this page should be re-examined for accuracy. I've come across several spells on a couple classes that seem to have incorrect coefficients posted. I'm sure that these were not inaccurate numbers at one time, but seeing as how spell damage coefficients are one of the most fluid mechanics in the game, many of them have been rendered inaccurate as the game has evolved.

Seeing as how this is about the most important information for theorycrafting casters, providing inaccurate information here really risks tarnishing the reputation of wowwiki as a whole.K10wn (talk) 17:25, 4 October 2008 (UTC)

Palading healing coefficients have been updated
Based on the numbers that I was provided from generous people and from my own math for my own guide, I updated the spellpower coefficients for the healing spells for paladins and added in Sacred Shield, because that spell is also affected by spellpower. I also adjusted the spellpower value of Holy Shock when it is used as an offensive spell. If the healing spellpower coefficients do seem high, keep in mind that it also incorporates the improved healing you get not only from talents but also from Blessing of Light being incorporated into the baseline healing abilities.Qygibo (talk) 19:22, 14 February 2009 (UTC)


 * Do you mean you didn't get the base coefficients? What were your test conditions? If you look at the rest of the page, coefficient-boosting talents' effects (e.g. Empowered Fire) are shown separate from the baseline (untalented, unmodified) coefficients. Effects like Darkness that don't directly boost the coefficient aren't even counted. Adesworth   talk to me IconSmall Mage.gif 19:45, 14 February 2009 (UTC)

After rereading, I was incorrect in stating that the spellpower coefficients posted were taking into account talents; they are base coefficients. These numbers are correct from various theorycrafters as taken from http://www.elitistjerks.com and from various other sites. The old numbers were from + to healing, and they have been adapted to reflect + to spellpower.Qygibo (talk) 20:57, 14 February 2009 (UTC)

Pet coefficients
This article could use spell and attack power coefficients for hunter and warloc pet abilities. -- Gordon Ecker (talk) 03:32, 15 February 2009 (UTC)
 * If you move the page we'd have to include attack power coefficients of Hunter, Rogue, Warrior, Shaman and Death Knight abilities. There's nothing too big about the page that justifies the move right now. It lists spell power coefficients, and Paladins are merely a special case. Adesworth   talk to me IconSmall Mage.gif 14:48, 15 February 2009 (UTC)
 * A separate attack power coefficients article could be created. Either way, I don't see why this article shouldn't have pets' spell power coefficients. -- Gordon Ecker (talk) 01:29, 17 February 2009 (UTC)
 * I've made one and linked to it. -- Gordon Ecker (talk) 01:37, 27 February 2009 (UTC)

Ability spam
I think it will be a good idear to spam ability tags onto the abilitys to make it easyer for people just looking at the page to see what the ability does. Richwarf (talk) 10:24, May 27, 2010 (UTC) (tag added a few days later)


 * All done, onto AP page :P Richwarf (talk) 13:33, May 30, 2010 (UTC)