The way the section is currently presented, the conclusion may lead some shaman to assume that Stonebreaker is only good if they can maintain constant shock spam. The math tells a different story, and I think it would be better to rephrase the conclusion and sample values used to more accurately reflect this.
I agree that this section is confusing as written. I took a different route to solving this problem, and determined that you need ~30-35% buff uptime for Stonebreaker's to be better than Astral Winds. Shock miss rate would obviously impact this buff uptime but it's not required to be considered as a separate variable unless we know the mechanic by which Stonebreaker's procs (PPM? Percent chance? Internal cooldown, if any?) Unfortunately, although I recall posting that analysis to EJ, and getting a response, I can't find my post, and I'm too lazy to re-do the math... feel free to search for it, but I think I must have been sleepy while posting, hit the wrong thread, and gotten my post deleted.
It's the miss rate of your SHOCK spells, doesn't have anything to do with melee hit rate.
I guess that's yet another unclear aspect of the section then, cause I obviously misinterpreted it as melee miss rate, not shock miss rate. Still, what kind of shock miss rates can we expect as enhance shaman? Surely it's much closer to 25% than 77%, no?
Edit: Anecdotally, WWS shows my personal shock miss rate to be somewhere in the ballpark of 10%.
Originally Posted by Rob
...unless we know the mechanic by which Stonebreaker's procs (PPM? Percent chance? Internal cooldown, if any?
You have 17% base chance to resist a spell vs a lvl 73 mob (of which 16% can be mitigated by talents/spell hit rating). With Nature's Guidance, you reduce your miss chance to 14% .
You have 17% base chance to resist a spell vs a lvl 73 mob (of which 16% can be mitigated by talents/spell hit rating). With Nature's Guidance, you reduce your miss chance to 14% .
Ok, so back to my original point then. A 14% miss rate on shocks leads to a required uptime of around 28%, so we're still in the general 30% uptime range.
Edit: Furthermore, the shock miss rate for enhancement shaman is going to be much more consistent from one person to the next than the melee miss rate, so it seems even more logical to use a realistic value here (i.e. 17-14% instead of 93% or 77%).
OK, here's a quick and dirty replication of my math. Last time I did it, I accounted for white misses but said that 50% of WF procs were MH and 50% were OH. This time, I'm too lazy. The net result of these two exclusions from the model is a higher required uptime. Real number should be closer to 30%.
In a 30 second period, we get at most 10 MH WF procs, 3 stormstrikes, 15 MH attacks, and 15 OH attacks (weapon speed = 2.6, 100% flurry uptime).
Stonebreaker applies to everything, but it's halved on the offhand.
10*2*110*1.4 + 18*110 + 18*110*0.5 = 3080+1980+990 = 6050 bonus AP.
Solve for uptime p such that
2240 = 6050p
p = 37%
Stonebreaker is therefore superior if uptime exceeds 37%.
Complicating factors not accounted for:
* Increased white miss/glance rate (vs. yellow)
* Increased critical strike chance of white attacks (vs. yellow)
* This is a best case scenario in WF procs (you will not get 10 MH procs)
* This is a best case scenario in Flurry uptime (you will not have 100%)
By far, the largest of these factors is the exceedingly small probability of getting exactly 10 MH WF procs in a 30 second period. Therefore, this is an overestimate of the required uptime for Stonebreaker to be superior.
It's the miss rate of your SHOCK spells, doesn't have anything to do with melee hit rate.
I'm quoting this and addressing it again, because looking at the numbers, I'm not convinced that it is spell miss rate and not melee miss rate being considered.
A: Astral Winds AEP = Percent_of_WF_Swings * Astral_Winds * Elemental_Weapons
B: Stonebreaker AEP = Uptime + B1 + B2 + B3 <====typo in the OP
B1: Stonebreaker AEP to WF = Percent_of_WF_Swings * Stonebreaker * Elemental_Weapons
B2: Stonebreaker AEP to SS = Percent_of_SS_Swings * Stonebreaker
B3: Stonebreaker AEP to white damage = Percent_of_white_Swings * Stonebreaker * (1 - MissRate)
If it's spell miss rate, why does it only appear in the white swings term, and not every term with the Stonebreaker ap? Isn't spell miss rate most easily taken into consideration as a reduction of the 50% proc chance on the effect?
Edit: there's also a typo in the OP, it's corrected later on in the same post. the line should read:
B: Stonebreaker AEP = Uptime * (B1 + B2 + B3)
OK, here's a quick and dirty replication of my math. Last time I did it, I accounted for white misses but said that 50% of WF procs were MH and 50% were OH. This time, I'm too lazy. The net result of these two exclusions from the model is a higher required uptime. Real number should be closer to 30%.
In a 30 second period, we get at most 10 MH WF procs, 3 stormstrikes, 15 MH attacks, and 15 OH attacks (weapon speed = 2.6, 100% flurry uptime).
Stonebreaker applies to everything, but it's halved on the offhand.
10*2*110*1.4 + 18*110 + 18*110*0.5 = 3080+1980+990 = 6050 bonus AP.
Solve for uptime p such that
2240 = 6050p
p = 37%
Stonebreaker is therefore superior if uptime exceeds 37%.
Complicating factors not accounted for:
* Increased white miss/glance rate (vs. yellow)
* Increased critical strike chance of white attacks (vs. yellow)
* This is a best case scenario in WF procs (you will not get 10 MH procs)
* This is a best case scenario in Flurry uptime (you will not have 100%)
By far, the largest of these factors is the exceedingly small probability of getting exactly 10 MH WF procs in a 30 second period. Therefore, this is an overestimate of the required uptime for Stonebreaker to be superior.
I'm not sure that halving all the bonus AP applied to OH attacks is the proper way to approach this. I'm having trouble showing myself exactly why I feel this way, but it seems a bit off.
On that note, why not split the difference and say 5 MH WF procs and 5 OH WF procs, in which case your method would require halving the bonus AP from Astral Winds on OH attacks. This in turn would lower the "total bonus ap" contribution from Astral Winds to 1680.
Either way, I think Panny's method from the OP is closer to reality.
Since the Stonebreaker Totem affects white damage where as Astral Winds doesn't, I needed to work out how much the contribution was. Therefore, melee miss rate affects the Stonebreaker Totems contribution and needed to be calculated. I am no longer sure about the hit cap and the miss rates of dual wielding Shaman, so I left it as a variable while writing out the equation.
After I did my experiments, melee hit rate (miss rate) was the only unknown I had left, so I did the logical thing and solved for it. Rob is correct when he mentions AP being halved on the offhand. However, since it affects both Astral Winds and Stonebreaker equally (1.5 multiplier on both), I didn't include it.
Alot of my calculations are relative and set based rather than absolute because there is alot I haven't verified for myself, not fully sure of and wasn't sure on how to include. (For example, WF proc rate while dual wielding, flurry uptime, hit caps, etc). In these cases, I used values that favoured the Astral Winds because figuring out whether Stonebreaker was an upgrade was my main goal. Figuring exactly how much better it was was more daunting and so I used fuzzy logic for this. I mentioned this in my original post concerning the factoring of spell resist rates (which is just taken as a simple multiplier.)
One thing I did forget to include was glancings and the different crit rates from yellow to white damage (yellow is on a two-roll system so the crit rate will be lower?). These can be pretty easily added if there are commonly agreed values for them.
B: Stonebreaker AEP = Uptime + B1 + B2 + B3 <====typo in the OP
That should be B: Stonebreaker AEP = Uptime * (B1 + B2 + B3)
Since the Stonebreaker Totem affects white damage where as Astral Winds doesn't, I needed to work out how much the contribution was. Therefore, melee miss rate affects the Stonebreaker Totems contribution and needed to be calculated. I am no longer sure about the hit cap and the miss rates of dual wielding Shaman, so I left it as a variable while writing out the equation.
After I did my experiments, melee hit rate (miss rate) was the only unknown I had left, so I did the logical thing and solved for it. Rob is correct when he mentions AP being halved on the offhand. However, since it affects both Astral Winds and Stonebreaker equally (1.5 multiplier on both), I didn't include it.
Alot of my calculations are relative and set based rather than absolute because there is alot I haven't verified for myself, not fully sure of and wasn't sure on how to include. (For example, WF proc rate while dual wielding, flurry uptime, hit caps, etc). In these cases, I used values that favoured the Astral Winds because figuring out whether Stonebreaker was an upgrade was my main goal. Figuring exactly how much better it was was more daunting and so I used fuzzy logic for this. I mentioned this in my original post concerning the factoring of spell resist rates (which is just taken as a simple multiplier.)
One thing I did forget to include was glancings and the different crit rates from yellow to white damage (yellow is on a two-roll system so the crit rate will be lower?). These can be pretty easily added if there are commonly agreed values for them.
That should be B: Stonebreaker AEP = Uptime * (B1 + B2 + B3)
In that case, I would change the analysis a bit. You can get the %WF, %SS, and %white from WWS logs easily enough. Taking these numbers from the largest sample of my own I have (a night in SSC) gives the following:
Physical Damage breakdown:
From WWS reports over the course of an entire raid night:
Percent_of_WF_Swings = 0.255
Percent_of_SS_Swings = 0.130
Percent_of_white_Swings = 0.615
As miss rate decreases, so does the required uptime to break even. An enhancement shaman with +9% hit from talents and no hit rating on gear would see a 19% miss rate against boss mobs, resulting in a required uptime of 26.4%.
Changing some values for more hit gear, we find that even if your miss rate is better than 2% the required uptime is 23.8%. This gives us a fairly good window of possible required uptimes for the Stonebreaker buff over a wide variety of gear, with a good conservative estimate being 30%.
Now what does it take to achieve a 30% uptime? Here’s a simplified formula for determining this.
(Seconds per Shock) = (Stonebreaker proc chance) * (1-Spell miss rate) * (Stonebreaker duration) * (1/Stonebreaker Uptime)
(Seconds per Shock) = (0.50)*(1-0.14)*(1/0.30) = 14.3
This means that on average, a shock every 14.3 seconds is the minimum needed to maintain 30% uptime on the totem buff and thus make it come out ahead of the Astral Winds totem.
Conclusion: The Stonebreaker totem should be superior to the Totem of Astral Winds on average if the shaman can shock at least once every 14 seconds.
Edit: In the end, even tweaking the numbers around between the percentage breakdown in the OP, various miss rates, and bad spell miss rates, I always come up with a required shock frequency greater than 12 seconds (i.e. two shock cooldowns). I think that's the important punchline to the whole analysis.
In that case, I would change the analysis a bit. You can get the %WF, %SS, and %white from WWS logs easily enough. Taking these numbers from the largest sample of my own I have (a night in SSC) gives the following:
Sure, that's definitely one method of performing the calculations and presenting the conclusions.
However, I've found my own swing percentages recorded by WWS to vary significantly each raid. They're affected by the levels of the mobs you're hitting (did you include trash?), buffs debuffs for each raid (LotP affecting crit rates, Imp. FF affecting hit, etc), travel time in between mobs/damage windows, specific mob mechanics, and just how alert you were. I'd prefer to work out calculations for the base case and modify as needed.
Also, my instinct is to avoid statistics as much as possible. :P
Originally Posted by Skiace
Edit: In the end, even tweaking the numbers around between the percentage breakdown in the OP, various miss rates, and bad spell miss rates, I always come up with a required shock frequency greater than 12 seconds (i.e. two shock cooldowns). I think that's the important punchline to the whole analysis.
If your methodology is sound, I think presenting it this way is very good.
Sure, that's definitely one method of performing the calculations and presenting the conclusions.
However, I've found my own swing percentages recorded by WWS to vary significantly each raid. They're affected by the levels of the mobs you're hitting (did you include trash?), buffs debuffs for each raid (LotP affecting crit rates, Imp. FF affecting hit, etc), travel time in between mobs/damage windows, specific mob mechanics, and just how alert you were. I'd prefer to work out calculations for the base case and modify as needed.
Also, my instinct is to avoid statistics as much as possible. :P
If your methodology is sound, I think presenting it this way is very good.
What I'd like to do is take your formula and try out basically the worst case scenario of numbers we can realistically use. This would be close to zero +hit gear and thus a high %WF, coupled with a pessimistic spell miss rate. If this still gives a required shock frequency in the 12-14 second range, I think that should be presented as the proof for the totem's superiority. Of course, eventually it would be great to see it incorporated into Yo's sim as well.
Does anyone have a swing breakdown favoring Windfury more than the one seen in the OP?
(that would be 28.8% WF, 11.9% SS, 59.5% white)
Also, if anyone disagrees with the method I used to determine shock frequency from needed buff uptime, please speak up.
I'm not sure that halving all the bonus AP applied to OH attacks is the proper way to approach this. I'm having trouble showing myself exactly why I feel this way, but it seems a bit off.
Well, I'm going to have to disagree with you here unless you can show me any reason at all for why you feel this is wrong -- remember that we daily have people feel that hit rating should be valued more highly. Offhand attacks do half damage. Therefore they're worth half as much bonus AP in my calculations.
On that note, why not split the difference and say 5 MH WF procs and 5 OH WF procs, in which case your method would require halving the bonus AP from Astral Winds on OH attacks. This in turn would lower the "total bonus ap" contribution from Astral Winds to 1680.
Yep, as I acknowledged, this is a "best case" for Astral Winds. What you suggest is perfectly reasonable, and I did it last time I did these calculations. I did not do it this time because I also did not factor in white misses or glances (a problem which yellow attacks do not have), and I also didn't factor in yellow attacks having ~5% lower crit rate, but it is trivial to adjust for this single factor if that's what you dislike about this analysis. As you said, Astral Winds becomes 1680 bonus AP. Stonebreaker's WF attacks are also lowered and its AP contribution becomes 10*2*110*1.4*.75 + 18*110 + 18*110*0.5 = 2310+1980+990 = 5280. Uptime required is p = 32%. Not an earth-shatteringly different result.
Either way, I think Panny's method from the OP is closer to reality.
Not sure how that is the case when you suggest that you must only shock once every 14 seconds to make Stonebreaker's superior (a result which absolutely will give sub-50% uptime) and the OP says that shock spam and 52% uptime is required to make Stonebreaker's superior.
While roaming WoW, I bumped into an enhancement shaman who claimed that STR had diminishing returns, therefore go AP. Now I know that this thread has always said STR > AP, but is there a point of diminishing returns? I know there are for crit, dodge, defense, etc... but I have never heard of one for a stat. Figured I would go to the pros and ask.
While roaming WoW, I bumped into an enhancement shaman who claimed that STR had diminishing returns, therefore go AP. Now I know that this thread has always said STR > AP, but is there a point of diminishing returns? I know there are for crit, dodge, defense, etc... but I have never heard of one for a stat. Figured I would go to the pros and ask.
Thats the dumbest thing i have ever heard. Str always is 2 ap (kings yadda yadda), and 1 ap is always 1 ap. The only thing i can think of is how, from what I've heard, when itemizing stuff blizz starts applying higher weights to items that have more of one stat (ie 10 str is less expensive than 20 str but is the same as 10 str on 2 diff items) in either case the same rule would apply to ap anyway and the point is moot even if thats what he thinks he talking about.
Sorry for the irrelevant matter, but i wanted to ask something. On the starter post of this thread, it is stated that shocks are 5-6% of total enhancement dps. Are you sure that this is right?
My calculations show that even before 2.3, and without talents at all, and with only earth shocks we have :
675 damage every 6 seconds= 112 dps. If we factor in 5% base spell crit and the resist chance on a level 73 boss we have about 100dps. Finally adding the SS debuff we go to 120 dps. Now in order for that to be 6% of total damage, your total damage must be. 120/0.06=2000 dps. Which is much higher than a normal raiding shaman can achieve.
So correct me if im wrong, but my numbers show that even in the worst occasions shocks are a lot more of shamans dps than 6%. So maybe we need to reconsider speccing elemental in patch 2.3, where the shock damage will increase greatly, while its cost will be almost free.
Either way, I think Panny's method from the OP is closer to reality.
Not sure how that is the case when you suggest that you must only shock once every 14 seconds to make Stonebreaker's superior (a result which absolutely will give sub-50% uptime) and the OP says that shock spam and 52% uptime is required to make Stonebreaker's superior.
The bolded claim demonstrates the exact the point I've been trying to make. The OP says that you need greater than 52% uptime IF your miss rate is 77%. My whole point originally was that this is a misleading way to present the data, and would lead to people coming to the conclusion you did here.
Using the same exact formula and numbers from the OP, (Panny's method is the same one from the OP that you referenced) but solving for uptime with a more realistic miss rate of 19%, I got a required uptime in the range of 30%, which is where I got the 14 seconds from.
edit: I'm also pretty sure that Thundering Strikes now properly applies to yellow attacks, so they do not see a 5% reduced crit rate. If you have a recent WWS that refutes this, please share.
Originally Posted by panny
However, I've found my own swing percentages recorded by WWS to vary significantly each raid. They're affected by the levels of the mobs you're hitting (did you include trash?), buffs debuffs for each raid (LotP affecting crit rates, Imp. FF affecting hit, etc), travel time in between mobs/damage windows, specific mob mechanics, and just how alert you were. I'd prefer to work out calculations for the base case and modify as needed.
I've found my swing percentages to vary no more than 3% in any given category between raids. This variation leads to a variation in the calculated uptime of less than 2%. If we can find some agreed upon base values I'd love to use them though. Following up on this, even if I bump up the WF swings to 30% (and reduce white swings accordingly) which is significantly higher than the OP values, I still end up with an uptime of 30% for a 19% missrate. I have to go all the way to 37% WF before the uptime dips below 12 seconds per shock. The more I look at it, the more I think 30% uptime is a good conservative estimate.
Sorry for the irrelevant matter, but i wanted to ask something. On the starter post of this thread, it is stated that shocks are 5-6% of total enhancement dps. Are you sure that this is right?
My calculations show that even before 2.3, and without talents at all, and with only earth shocks we have :
675 damage every 6 seconds= 112 dps. If we factor in 5% base spell crit and the resist chance on a level 73 boss we have about 100dps. Finally adding the SS debuff we go to 120 dps. Now in order for that to be 6% of total damage, your total damage must be. 120/0.06=2000 dps. Which is much higher than a normal raiding shaman can achieve.
So correct me if im wrong, but my numbers show that even in the worst occasions shocks are a lot more of shamans dps than 6%. So maybe we need to reconsider speccing elemental in patch 2.3, where the shock damage will increase greatly, while its cost will be almost free.
I think shocks are closer to 8-9% in the best case scenario. That's not a whole lot. I average about 1350 dps through all of BT/Hyjal, and on the melee friendly fights between 1650-1850 dps. That number can definitely drop between 5-6%, and if you're not shocking every 6 seconds on the dot, it'll drop even lower. Regardless, elemental is definitely becoming more viable.
If Totemic Mastery is unnecessary in the Sunwell I might be tempted to try it out. Mana is going to be such a non-issue, I will be casting a shock and fire nova totem every cooldown for sure. At that point, talents like Concussion, Reverberation and Call of Flame might actually beat out Nature's Guidance. I doubt it, though. Those talents really are pretty terrible for PVE.
If you're only using one axe, it should in theory only reflect the dodge rate for that weapon, so for SS purposes you'd want the axe in your MH to get that little extra edge. Kind of depends on how they implement it though.
Actually, it looks like the orc racial applies to both weapons even if only one of your weapons is an axe. With a RT/Syphon combo and no expertise gear, my character sheet shows that I have 5 expertise, with no mention of the bonus being axe-specific. In the 2.3 changes thread, human rogues were reporting that a sword MH/mace OH gave them 10 expertise total.
Could just be a paperdoll error, and it's going to be a pain in the ass to test. For now though it looks like orcs will only need 1 axe to get full benefit of the racial, and it doesn't matter which hand the axe is in.
I don't think the paperdoll is supporting weapon skill specific expertise.
I think the expertise part for axe is handled purely on server, since it basically is not supported designwise. Atleast imho the whole expertise change is to simplify the whole damage calculation, since you basically save a shitload of calculations for max hit cap etc. and you can technically reuse the attack table for both hands.
As you know I do not often post here, but something has been puzzling me for sometime and I think I may be making sense of it so here goes. I am hoping you guys can help make some sense of what I have discovered from WWS parses.
Like all Shaman here I have used the advice and simulators to make informed decisions about gear combinations to maximise my DPS in raids. In my case the simulators always seem to give me higher DPS potential than I actually achieve. I have put this down to the variations in bosses and have really only foccussed on Bosses such as Teron (the more static bosses) to get a good feel for whether I am hitting my potential.
I think like all of us here for whatever reason it matters to be always doing your best and pushing the limits. I live in 90-100% threat land and I love it.
But....I have looked at lots of WWS parses over the months and seen many Enhancement Shaman with better DPS than mine even when I would question the gear selection and or wether the gear is even better than mine. I go to the armoury do a comparison and usually I find I have better melee stats all round or at least comparable. So i kept asking myself what can I do better where am I going wrong....you know like you do :P
I have followed a number of enhancement Shaman as I have progressed through BT and Hyjal, Seb, Stig, Tsalrioth and Grevan. (Stalking the Armoury so to speak)
So after a long intro :-) today I started looking at WWS parses and found some differences I don't understand, but lead me to believe that chasing DPS is not all its cut out to be. In fact I would suggest damage done is important to.
Since most fights are limited by a bosses HP (even ones that heal up etc) we can look at Time to death as being variable. (yes this is obvious I know) and damge need to kill the boss pretty much fixed, if it does increase then so does time to kill. Without exception the highest dps kills on WWS appear to me to be those done in the shortest time. So you see top rating WWS parses for bosses from the likes of Blood Legion and Juggernaut etc. with huge dps values.
What I did was look at enhancement Shaman in these groups and looked at their DPS, discovering that they were ususally more than my DPS for the same boss with similar gear. Some foccused more on haste than me and others more leather...but our melee stats are fairly similar, weapons the same. Closer investigation showed that most but not all are advocates of twisting, either WF/GOA or WF/TA. I do not do this often so in the parses I usually see they use fewer or no shocks whilst my usage is higher.
It was at this point I decided that there must be another reason to account for such large differences. So when looking at Damage Done i started to notice some clear differences. Boss for boss on average I was doing a lot more damage than other Shaman but getting lower DPS figures. But their kills where quicker, their raid make ups different (usually more focussed on melee).
So in summary: I discovered I was doing more damage over a longer period with a lower dps and others are doing more dps over a shorter period. This of course all makes mathematical sense at least to me. I suppose time to kill is a combination of the sum of the parts in terms of DPS potential...certain mixes will get things done quicker and DPS per individual will be higher.
Finally to conclude, does my logic make sense or am I stating the obvious? Have the stats got the better of me? Does it really matter :-) ? Just really wanted to share this as dps is just a representation of a combination of stats as of course is damage done. Its just what you interpret from them that matters. I felt that I was doing something wrong with my gear selection etc and chased around for ages trying different things.
I'm not sure I fully understand what you'r saying, but I'll give it a shot:
The sim calculated your possible best DPS. Due to the nature of shamans (SR), this DPS should be pretty much sustainable over a longer period of time.
Because your kills take so much longer, you will have more damage done then the shaman you linked. The main difference here is that your group does less damage then their group.
Say you have a mob with 100k hp and 2 dpsers.
In your case:
you - 500 DPS --> 55555 damage
partner - 400 DPS --> 44444 damage
In their case:
Shaman - 700 DPS --> 42424 damage (less then you)
Partner - 950 DPS --> 57575 damage
To clarify: DPS is not -or should not be- influenced by time, but damage done depends on the DPS of other people aswell. With a raid that does significant more DPS, your contribution in that DPS might be lower then in a raid that does less DPS.
In your raid, avg dps per person (including healers) is 667, you almost double that number.
In Blood Legion, avg dps per person is 1065.48, doubling that as shaman is going to be alot harder.
@Everyone talking about the Stonebreaker totem - all I did was copy Panny's derivations. You guys come up with something agreed on that represents it better, I'll update that section.
@Lujaar - that same thread points out that the paper doll is obviously not reflecting how the expertise actually works. The paper doll is just going to sum the expertise, the mechanics in the background are going to actually split the expertise between general expertise from gear, and the racial expertise.
@Boro - comparing WWS between guilds is incredibly difficult. There's a ton of factors that could be involved as far as what debuffs were on the boss, how aggressive they were in pushing the threat window, whether they were using Haste pots and getting Drums, among other things. Twisting as I've found out is incredibly overpowered for one thing - GoA totem is worth a hell of a lot more than shocks according to my WWS parses where I've used it.
Regarding shock % of dps - I just went through 15 WWS parses of top guilds and not a single Enhance Shaman had more than 4% of their damage done from shocks. When I calculated how many shocks were done versus the length of the fight, most of the shamans had about a 20-30 seconds period of no shocks done (warmup period for threat most likely) so they could have done a little bit more, but it certainly would not be pushing much higher than 5-6%.
11/07/07, 8:33 PM
Similar Threads
