The one thing fire gives is -35% cost for blast wave / flame strike and higher aoe burst in form of combustion > scorch for 2 crits and Flamestrike > BW for 2x damage.
A little bit more to our Raid Constellation: We got a fix party with 2-3 Mages and 1-2 Shadow Priests. very few times a shaman or random caster/pala. We saw that for our setup fire specced Mages fit best and maybe it´s kinda surprising or not the mages often do the most dmg at all together with the rogues.
That´s the most important reason for me to not try out ice specc at the moment. At the moment all 3 mages got imp scorch, so we got no problems keeping the stack of scorch up, even with 2 mages its no problem.
If i take the german inwow theory for the crit/hit/dmg calculation, i would say, I should now focus more and more on spelldmg. Another question which came in to my mind is, is there a fix hit cap, which I should focus on, yet I got 5% equip based and 3% via icetalents, but i couldn´t really find out somewhere if 8% is enough if you put some more other stats on the equip or if I maybe should focus 10%.
Alright, back to the Frost vs Fire resist for the Eye of Mag.
Faxmonkey had posted up that the frost gets more from the Eye because frost spells are binary and actually have a 3% innate resist, as opposed to Fire's 1%. Or something to that effect. Here's the quote:
". It will proc constantly with a frost build. It's a ridiculously good trinket and you must want to have it, even if you have good +hit and have a fire spec, I'm pretty sure that there's an unmitigatable 3% chance to resist versus a skull mob (and 1% on everything else). With just that 3% this trinket easily outperforms most other DPS trinkets and that's before the buff it's receiving in 2.1."
Which caused my eyes to pop out, considering i've always read 1% for both fire and frost.
Several people called BS, others agreed:
"Also, there isn't really a 3% miss cap vs skulls (I bet whats causing you to see that is the unmitigatable ~15-30 resist a skull mob has due to level and that frostbolt is binary - while its probably true for frost its not true in general)."
and
" To elaborate, it's 3% for frost because monsters gain +5 resistance than cannot be reduced for every level above you they are. Thus, frost has 15 unavoidable frost resist on lvl 73 bosses, which increases the binary resist rate, plus the minimum 1% miss chance."
Now slap me silly, but if the above is true, then it means two things: The eye is much better for a frost mage and two, a frost mage only needs +13 ~ 14% hit, instead of the regular 16% to hit??
I assume since they mention the 15 level resist can not be mitigated by spell penetration?
Or should I just be slapped for reading the wow mage forum?
People are being confused by the "glancing blow" mechanic that non-binary spells suffer from (aproximately 5% damage loss due to partial resists IIRC), adding it on to binary spells, and are then coming to a false conclusion.
Non-binary spells have a glancing blow mechanic of partial resists that can not be overcome.
Binary spells don't.
As a side note, Eye of Magtheridon did not proc from a resisted Impact.
" To elaborate, it's 3% for frost because monsters gain +5 resistance than cannot be reduced for every level above you they are. Thus, frost has 15 unavoidable frost resist on lvl 73 bosses, which increases the binary resist rate, plus the minimum 1% miss chance."
This is true, and is very easy to check.
Get someone in frost resistance gear (from sapphiron times) and spam frostbolts on him. You will notice lots of resists even if you have capped hit vs lvl 70. That means that resistance works against binary spells as well. It is harder to prove that +hit doesnt prevent resists caused by resistance on binary spells, as you will need lots of samples to notice difference in amount of resists when using gear with low and high spell hit rating. Logically it shouldnt, because it doesnt for fire.
However, i dont think that you need less hit% if you are frost specced. I cant prove it, but it would be logical if those resists caused by targets resistance added to "miss" type resists, caused by level difference.
This is true, and is very easy to check.
Get someone in frost resistance gear (from sapphiron times) and spam frostbolts on him. You will notice lots of resists even if you have capped hit vs lvl 70.
PvE and PvP hit mechanics are different, I wouldn't go that route for testing.
This is true, and is very easy to check.
Get someone in frost resistance gear (from sapphiron times) and spam frostbolts on him. You will notice lots of resists even if you have capped hit vs lvl 70. That means that resistance works against binary spells as well. It is harder to prove that +hit doesnt prevent resists caused by resistance on binary spells, as you will need lots of samples to notice difference in amount of resists when using gear with low and high spell hit rating. Logically it shouldnt, because it doesnt for fire.
However, i dont think that you need less hit% if you are frost specced. I cant prove it, but it would be logical if those resists caused by targets resistance added to "miss" type resists, caused by level difference.
Logic has nothing to do with it.
It's been proven that a hit cap frost mage and a hit cap arcane mage will HIT 99% of the time on a +3 level target. The arcane mage will lose a percentage of their damage due to partial resists, while the frost mage won't. Go a couple pages back in this thread and look at Yeonora's data.
---
As far as resisting when the target has a real resist score to overcome, +hit% can overcome that for binary spells but won't for non-binary. The resist check for binary spells is thrown into the hit check. The resist check for non-binary spells is checked on damage (a seperate roll from the hit check), where the "glancing blow" mechanic is also applied.
Merith it's very well known that +hit does not counteract resists from a Resistance value greater than zero. That's what penetration is for.
The issue of debate is what Resist mobs get for having a higher level than you. The evidence seems to point to some amount of partial resists for non-binary and no penalty for binary. But I have a hard time believing this is true. I think it's more along the lines of the partial resist is easier to notice in a smaller sample size because it can apply itself in a small way (25% resisted) whereas the only way it can take effect on a binary spell is a full resist, and those would have to be only 1 in 20 if we assume it's a 5% damage penalty. And it may be something as small as 3%.
In either case 3%/5% is such a low proc rate that using Magtheridon's over Quagmirran's, the Icon and/or Neltharion's Tear seems suspect to me. It can't proc off partial resists, so that mechanic would only help Magtheridon's Eye if the penalty applies to binary spells. *If* it does, and *if* we assume the 5%, that's one proc every 50 seconds. That's more than twice as often as the Icon but you do not get to choose when it happens. You can't line it up with Arcane Power, Bloodlust, etc. And the Icon's buff lasts twice as long.
The massive +54 damage cannot be ignored, it may be worth wearing for that after the patch. But I fall firmly into "That proc is poorly designed garbage." camp.
There is absolutely no proof that binary spells are immune to the level based resists at this point.
The only data we have seen is with a sample size of ~350. That is way way to low to be of any statistical significance. The person testing this said s/he would collect more data this weekend and report back.
If binary spells do infact ignore level based resists than it is something introduced in TBC as it was certainly not the case pre TBC.
To kind of clear this mess up a bit. To my understanding:
- Binary spells do not suffer from the 5-6% resists against mobs three levels higher than you.
- Non-binary spells do suffer from 5-6% resists against mobs three levels higher than you in partial resists.
- You cannot overcome this "glancing" mechanic with curse of shadows/elements or spell penetration on gear.
- You cannot overcome "real" resistances (for example vs. netherdrakes or lvl70 players with frost resistance in PvP) with more hit rating on gear. Only spell penetration can do this.
- With 16% hit total, you get in average 1% complete resists against every mob three levels higher than you with every spell, provided the mob does not have additional "real" resistance thrown on that you do not penetrate.
- This has been exactly this way since Blizzard implemented the patch that changed negative resistances (through curses for example) to not longer produce negatively resisted spell hits that are way larger, but instead threw 10% additional spell damage done on curse of the elements and curse of shadows. I experienced this same behaviour in AQ40 and Naxxramas, I was frost, arcane/frost and deep fire specced, and I looked it up in swstats.
The problems: I do not have any data saved from ye olden times that would prove my point. Edit: But we do have Kavan's data (see two posts below).
I (we) do not have more than 730 samples for frostbolt that I collected. Out of those samples (349 on Kharazan bosses, 381 on Magtheridon and his channelers), 7 have been resisted (0.96%). I was running around with almost exactly 16% hit for those samples.
I do not have more than 655 samples for arcane blast. Out of those samples (316 on Kharazan bosses, 339 on Magtheridon and his channelers), 7 have been resisted (1.07%), 2.8% on Kharazan bosses has been partially resisted (but as we noticed, you do, strangely enough, never get partial resists on Netherspite), 3.9% on Magtheridon and channelers has been partially resisted.
Kavan should have lots and lots of data for arcane spells in his ResistTracker, but split up into bosses, hit rating, extra hit, spell penetration.
I don't remember enough of the statistics stuff I learned 5-10 years ago to calculate how many samples we need to say: "Well, with just a 5% error margin, this data is significant and shows that you just get 1% resists with frostbolt instead of 3%" (or whatever some other guys try to tell us). Maybe some of you (Kalman?) could tell us how many samples we need.
If any of you has a great idea how to do a longer test series, please tell me. I obviously can't convince my raid to do 20 Gruul tries just standing there watching me spam frostbolt. Did any of you try Anachronos and knows if he does have glancing mechanics or not?
Another open question you can help us with: Do ignites double-dip on resists? Meaning, can the ignite ticks themselves be resisted? One screenshot of the crit partially resisted and then an ignite tick partially resisted would be enough. If they do, fire loses some more damage due to level-based unmitigable resistances.
Now, the data from this week's Gruul and Magtheridon kill is very unsatisfying. We downed Gruul the first attempt (meaning very few samples). We wiped 5 times to channelers and downed Magtheridon the first attempt where we finally managed to get into phase 2 clean. But me and our only druid have been killed by infernals in phase 1 (meaning very few samples). Plus the elemental shaman in my group died very fast on some of the magtheridon attempts, so I only had 12% hit for a significant amount of time. Which kind of renders this week's crop useless :-/
- You cannot overcome "real" resistances (for example vs. netherdrakes or lvl70 players with frost resistance in PvP) with more hit rating on gear. Only spell penetration can do this.
Are you sure about this? Last I checked, I was able to use hit gear to overcome resistance on binary spells. That was before Naxx and I didn't test for very long.
Spell penetration is the superior option anyways, unless spell hit can somehow overcome stuff like Improved Concentration Aura.
As far as I'm aware spell penetration affects neither the "resist pushback" portion of Concentration Aura or the "resist silence" portion of Improved Concentration Aura.
I'm trying to quantify the actual loss of dps from level-based partial resists.
I selected random anonymous logs on the WowWebStats server, and checked how their warlocks and mages performed. WowWebStats gives the number of DoT ticks / spells landed and the average mitigation due to partial resists. I used logs from Magtheridon (the adds are 73) and Gruul.
I mixed up everything that had more than 40 occurences: DoTs, SBs, Fireballs, scorches (except Immolate, because it has a strange mix of DD and dot, and Siphon Life, because it's binary).
Taking 68 different series of spells, the average amount mitigated is 4.257 %. That same average, but weighted by # of ticks/hits, gives 4,344%. Standard error (between series of spells, unweighted) is 1,435%. The total number of ticks/hits/crits in my sample is 4652.
Warlocks had an average loss .1-.2% higher than mages. The difference probably isn't statistically significant, however.
This experiment isn't completely satisfactory, since I wasn't exactly in a position to control the environment much, but it does point out that the 6% loss from partial resists calculated pre-BC no longer applies.
Oh, and to answer your question, Yeonora, in all the reports I checked, no Ignite had a single partial resist.
Last edited by Setia : 05/13/07 at 5:39 AM.
Reason: Spelling mistake.
You'll see that their reported mitigation is 19.0 %. This is they sum all values in (x resisted) and divide by damage done, which gives (360 * 10) / (1135 * 10 + 1513 * 5) = 19.0%.
This is wrong. What they should have done if they do it this way is compute percent damage resisted out of all possible damage, so (360 * 10) / (360 * 10 + 1135 * 10 + 1513 * 5) = 15.9%.
Now even this is wrong because it doesn't account for how combat log displays amplifiers. So the real %dmg lost to partial resists in this case is (360 * 10 * 1.05) / (1513 * 15) = 16.6%.
So they have it wrong on two parts, both of which my mod accounts for correctly having real time debuff/buff information about multipliers and knowing individual resists can only be in increments of 25%.
Let me give you another example since the above might make someone think it can only show higher number than it really is. The following values are taken from a real log (timestamps are not real)
5/10 21:34:14.078 Your Arcane Blast hits Hellfire Warder for 1629 Arcane damage.
5/10 21:34:15.078 Your Arcane Blast hits Hellfire Warder for 1721 Arcane damage.
5/10 21:34:16.078 Your Arcane Blast hits Hellfire Warder for 1619 Arcane damage.
5/10 21:34:17.078 Your Arcane Blast crits Hellfire Warder for 3681 Arcane damage.
5/10 21:34:18.078 Your Arcane Blast crits Hellfire Warder for 3715 Arcane damage.
5/10 21:34:19.078 Your Arcane Blast crits Hellfire Warder for 3913 Arcane damage.
5/10 21:34:20.078 Your Arcane Blast hits Hellfire Warder for 2370 Arcane damage.
5/10 21:34:21.078 Your Arcane Blast crits Hellfire Warder for 3244 Arcane damage. (588 resisted)
5/10 21:34:22.078 Your Arcane Blast hits Hellfire Warder for 2353 Arcane damage.
5/10 21:34:23.078 Your Arcane Blast hits Hellfire Warder for 1150 Arcane damage. (365 resisted)
5/10 21:34:24.078 Your Arcane Blast hits Hellfire Warder for 1499 Arcane damage.
5/10 21:34:25.078 Your Arcane Blast hits Hellfire Warder for 1554 Arcane damage.
5/10 21:34:26.078 Your Arcane Blast crits Hellfire Warder for 1365 Arcane damage. (743 resisted)
5/10 21:34:27.078 Your Arcane Blast hits Hellfire Warder for 1622 Arcane damage.
5/10 21:34:28.078 Your Arcane Blast hits Hellfire Warder for 1658 Arcane damage.
5/10 21:34:29.078 Your Arcane Blast hits Hellfire Warder for 1654 Arcane damage.
5/10 21:34:30.078 Your Arcane Blast hits Hellfire Warder for 1690 Arcane damage.
5/10 21:34:31.078 Your Arcane Blast crits Hellfire Warder for 2729 Arcane damage.
5/10 21:34:32.078 Your Arcane Blast crits Hellfire Warder for 2723 Arcane damage.
5/10 21:34:33.078 Your Arcane Blast hits Hellfire Warder for 1719 Arcane damage.
5/10 21:34:34.078 Your Arcane Blast hits Hellfire Warder for 1757 Arcane damage.
5/10 21:34:35.078 Your Arcane Blast hits Hellfire Warder for 1738 Arcane damage.
5/10 21:34:36.078 Your Arcane Blast hits Hellfire Warder for 1791 Arcane damage.
They report a value of 3.4% resisted (you can verify it is indeed the above mentioned procedure by summing resisted values and dividing by damage done).
In reality we have 2 25% resists and 1 50% resist, giving a (2 * 0.25 + 1 * 0.5) / 23 = 4.3%.
I hope this gives you some faith in the numbers I'm reporting.
Last edited by Kavan : 05/13/07 at 7:44 PM.
Reason: another example
You cannot overcome "real" resistances (for example vs. netherdrakes or lvl70 players with frost resistance in PvP) with more hit rating on gear. Only spell penetration can do this.
According to Blizzard you can. Can't find the archived post right now explaining how spell hit works, but basically as binary spells like frostbolt only have one resist roll, there is no difference between level based resists and resistance based resists for them - + to spell hit can and will lower the effect of the targets resistances.
This only works for binary spells though.
According to Blizzard you can. Can't find the archived post right now explaining how spell hit works, but basically as binary spells like frostbolt only have one resist roll, there is no difference between level based resists and resistance based resists for them - + to spell hit can and will lower the effect of the targets resistances.
This only works for binary spells though.
If you're thinking of this one, it doesn't tell us much, but to me it looks like Eyonix tells us that you cannot increase hit chance with more hit when a binary spell would be resisted because of resistances.
Eyonix the Mage (level 60) fires a frost bolt at Yeti of Doom (level 63). Eyonix is also wearing a total of +6% spell hit gear. Yeti of Doom has frost resistance such that he takes 50% from level 60 frost attacks. So, heres the hit calculation:
The game will roll a number between 0 and 1, and if its less than 0.445, the frost bolt will hit for full damage. Otherwise, a resist message will appear.
So basically you can bring the first part of the calculation to 0.99, but the second will always be 0.99*0.5 = 0.495.
I wouldn't trust WWS reports on partial resist information. Let me show you an example of how they compute the percentage.
(...)
I hope this gives you some faith in the numbers I'm reporting.
Thanks. So, what would be the post-BC number? Do mobs still have 8 spell resist per level, for a total of 24 ( = 5,14% partial resists according to the formula on http://www.wowwiki.com/Formulas:Magical_resistance), or is the SR adjusted for the mitigation to stay at 6%, as it was at 60?
Thanks. So, what would be the post-BC number? Do mobs still have 8 spell resist per level, for a total of 24 ( = 5,14% partial resists according to the formula on http://www.wowwiki.com/Formulas:Magical_resistance), or is the SR adjusted for the mitigation to stay at 6%, as it was at 60?
Search a few pages back for concrete numbers. From what I've seen my current theory is 6%.
I remember last year playing around with one of the spreadsheets that gave the relative value of the various trinkets. Vontre's spreadsheet gives us the ability to plug in upgrade values which is awesome.
Has anybody sat down and figured out the value of these procs with respect to gear and talents?
I was specifically thinking about Shiffar's Nexus-Horn.
The actual proc:
"Increases damage and healing done by magical spells and effects by up to 225 for 15 sec." which has a proc rate of 20% with an internal cooldown of somewhere of 30 to 45 seconds.
Summary: Fight lasted 11 minutes and 18 seconds including the 2 minutes which only the Channelers where engaged.
The trinket procced for 7 times. The time the proc was up (since it can not overlap due to the internal CD) was 105 seconds.
During the fight, I was only DPSing 85% of the time according to WWS due to running around for clicking on cubes.
I would like some help on figuring out the Theroycraft value of the trinket's proc.
I took some time to dig through the combatlog file, and the actual value of the proc was:
4,747 damage. This does not take into account Mag's vulnerablilty, CoE, Scorch Debuff or Molten Fury. I basically stripped out my damage when the Proc was up and assigned 225 for a fireball hit, 315 for a fireball crit (225+(225*0.4)), 112 for a Scorch hit and 157.5 for a Scorch crit (225*0.5+(225*0.4*0.5)).
I know I'm short changing the trinket by a good bit by not including the things like Scorch Debuff and CoE.
You'll see that their reported mitigation is 19.0 %. This is they sum all values in (x resisted) and divide by damage done, which gives (360 * 10) / (1135 * 10 + 1513 * 5) = 19.0%.
This is wrong. What they should have done if they do it this way is compute percent damage resisted out of all possible damage, so (360 * 10) / (360 * 10 + 1135 * 10 + 1513 * 5) = 15.9%.
Now even this is wrong because it doesn't account for how combat log displays amplifiers. So the real %dmg lost to partial resists in this case is (360 * 10 * 1.05) / (1513 * 15) = 16.6%.
Hullo,
The 19% mitigation was clearly a WWS bug, which is corrected now. If you go to the report, it now displays 15.9%, according to the second way of computing.
I don't understand the way you compute the third number. Where does this 1.05 factor comes from ?
Rest assured that I'm constantly tracking down bugs and miscalculations in WWS, and I'll always be glad to make corrections where they're needed. However, I'm alone on the project, and I might miss some bugs, so please do not hesitate to let me know, either on this forum, or, better, on the WWS forum here http://www.lossendil.com/forum/index.php
Last edited by lossendil : 05/19/07 at 12:10 PM.
Reason: Removing signature
The 19% mitigation was clearly a WWS bug, which is corrected now. If you go to the report, it now displays 15.9%, according to the second way of computing.
I don't understand the way you compute the third number. Where does this 1.05 factor comes from ?
Rest assured that I'm constantly tracking down bugs and miscalculations in WWS, and I'll always be glad to make corrections where they're needed. However, I'm alone on the project, and I might miss some bugs, so please do not hesitate to let me know, either on this forum, or, better, on the WWS forum here http://www.lossendil.com/forum/index.php
Cheers,
Lossendil
The 1.05 is likely to be misery. Since the Combatlog reports resistances before amplification you need to work out how much damage the resisted portion would have done. I'm a bit confused why curse of shadows 1.1 wasn't included though.
05/11/07, 3:52 AM
Similar Threads
