Tuskarr's Vitality should be classified as BiS. Are we not as realistic as possible with all?
By given simulation_time of 1000 the DPS differs a lot. 5000 hours is a more stable value.
Nevertheless I have done it to 7318 DPS with a few adjustments.
Thanks for that Rouncer. I tested as you suggest and yeilded the results you predicted. I will re-inlcude MW3_CL, MW4_LB in the ability priority order list.
Dedmon, I would not be willing to maintain such a page at this time as Wowhead is not able to list EVERY item that we would include, where as my html table can (though sometimes crudely, as is currently the case with Comet's Trail).
Dedmon, I would not be willing to maintain such a page at this time as Wowhead is not able to list EVERY item that we would include, where as my html table can (though sometimes crudely, as is currently the case with Comet's Trail).
Do we have a standard acceptable spec yet? The one I originally ran my tests with was without SF, with maxed out ImpSS, and nothing in Reverb. Is there a specific spec you want us testing with or are we trying to find the best spec as well? I've tested just about every variation of the spec I run with, but for some consistency a set spec would be nice.
Using what I believe is a fairly standard spec (linked below) I've pulled the following output from the gear listed on the first post.
After some confusion on my part, we clarfied simulation behaviour re: swing timer resets and casts. The result is that the sim is accurate in its recreation of the ingame environment, so the MW3_CL, MW4_LB could be re-included in results.
This put Aoelian's proposed set-up at roughly 7387 DPS. Duscha then provided a build that provided a stable 7390 DPS when I applied the rotation changes.
Duscha also raised a point about the stability of our results when using only 1000 hour sims. I think he is correct, and so have set down 5000 hours as the required sim time. This will limit the output range from 1-5 dps down to 1-2 dps.
Coming back to the issue of Cat's on boots, I am still not willing to apply it since a) its value is not equal in all encounters and b) our simulation tool cannot take these variances into account and c) it is possible for a raider to have 2 pairs of the same boots and apply different enchants to them based on encounter. At least this way, people will have a clear indication of theoretical maxims, while also understanding that Cat's will often provide superior DPS based on the amount of movement in the encounter.
As for spec, the inclusion of mana simulation as a compulsory element means that mana talents must be able to support the gear and rotation you have designed, otherwise the report would demonstrate a DPS decrease.
I have read through the Glyphs Numeric Comparison post and I dont see where the Windfury glyph has been proven to be surperior since the nerf from 5% to 2%. I dont have enhsim in front of me due to being at the office but does anyone have the comparison numbers with this gear set handy?
The last time I did a target dummy of like 4k swings I know the lightning shield glyph seemed to be like 30dps higher over the 3 or 4 rounds I did.
I have read through the Glyphs Numeric Comparison post and I dont see where the Windfury glyph has been proven to be surperior since the nerf from 5% to 2%. I dont have enhsim in front of me due to being at the office but does anyone have the comparison numbers with this gear set handy?
The last time I did a target dummy of like 4k swings I know the lightning shield glyph seemed to be like 30dps higher over the 3 or 4 rounds I did.
Target Dummy =/ Sim unless you managed to get an entire raid's worth of buffs and debuffs applied to the boss dummy.
My sim runs show WF > FT > LB for that third glyph spot.
Boot Enchants - I suggest removing them entirely for testing purposes.
In 3.2 there will be a third option that will muck up the BiS even further since Nitro boots will be getting 24 crit rating. Beyond that if someone is using two sets of BiS boots to be able to switch enchants based on the encounter then they will need to be hit capped in their cat speed enchanted version meaning Icewalker will take them over the spell hit cap. Since your current BiS uses Icewalker to just hit that spell hit cap it loses that ability to switch enchants, meaning it is no longer a real BiS set.
1. Major notice - it should be named as "[Enhancement] Enhsim BiS gear and set-up". There are other tools available such as Simcraft and Rawr and it is not proven that Enhsim is better reflecting in-game behavior.
2. Minor notice - why profession choice is limited to static 64 ap bonus? Jc+bs combo would add more flexibility in playing around hit/exp caps.
Again on Cat's on boots, I am still not convinced since there are at least two encounters currently in the game where this enchant provides no benefit whatsoever (XT normal mode and Patchwerk). Given that encounters such as these (stationary, tank and spank encounters) are often used as the 'benchmark' DPS encounters, they offer the greatest opportunity to view our maximum potential DPS in-game. I think there is some value in our simulation results, that demonstrate our theoretical potential, aligning with these in-game encounters that show our actual performance under conditions as close to the simulation environment as possible.
I make a clear note on the boots enchant directing people to the TTT section on minor speed to boots, This will allow people to make an informed choice about their boots enchant. If people do chose a run speed enchant for the currently listed BiS set-up, they will simply have to gear/enchant to accomodate this.
On professions, the JC, BS and Skinning provide non-uniform benfits that are not aligned with the standard +64 AP provided by most other professions. Since, as Rouncer points out, this is about gear and not professions, I actively sought to exclude professions such as these as they would skew the results, and make them impractical given that few people would spend 3-5K gold levelling a profession for a 10-15 DPS in simulator results. Any DPS increase that people of those professions are able to obtain is free DPS over and above the minimum laid down here.
As for the integrity of the Enhsim versus other sims, I will only say that Enhsim is as accurate a sim tool as the community is currently able to maintain. It provides a largely accurate testing environment and given the support provided by many people, bugs are often identified and corrected (such as currently occuring with the matter of swing timers and non-instant casts). Simcraft, according to one of the authors, is providing results with around 1% standard deviation from Enhsim, so I am sure that this is also a valuable tool. Rawr is not a useful tool for us as its since you cannot dictate an ability priority list.
Hey, my Chars name is Donodan, currently playing on Eu nathreuim as the raidlead form"set sail for fail" ~ world rank 190, therefore I am not new to the Ej forum.
Beacuse i am from Germany, my english isn´t the best , so I hope that you will forgive me for my wired structure and choice of words.
I was able to sim a log in which i reached ~7427-7430 dps
Hey, my Chars name is Donodan, currently playing on Eu nathreuim as the raidlead form"set sail for fail" ~ world rank 190, therefore I am not new to the Ej forum.
race orc
First off use [ code ] [ /code ] when posting a config file. Next, read the directions and understand them before posting. All sims are to be done with Tauren as the selected race to prevent race from being involved in the performance.
Originally Posted by Vespasian
Again on Cat's on boots, I am still not convinced since there are at least two encounters currently in the game where this enchant provides no benefit whatsoever (XT normal mode and Patchwerk). Given that encounters such as these (stationary, tank and spank encounters) are often used as the 'benchmark' DPS encounters, they offer the greatest opportunity to view our maximum potential DPS in-game. I think there is some value in our simulation results, that demonstrate our theoretical potential, aligning with these in-game encounters that show our actual performance under conditions as close to the simulation environment as possible.
I make a clear note on the boots enchant directing people to the TTT section on minor speed to boots, This will allow people to make an informed choice about their boots enchant. If people do chose a run speed enchant for the currently listed BiS set-up, they will simply have to gear/enchant to accomodate this.
On professions, the JC, BS and Skinning provide non-uniform benfits that are not aligned with the standard +64 AP provided by most other professions. Since, as Rouncer points out, this is about gear and not professions, I actively sought to exclude professions such as these as they would skew the results, and make them impractical given that few people would spend 3-5K gold levelling a profession for a 10-15 DPS in simulator results. Any DPS increase that people of those professions are able to obtain is free DPS over and above the minimum laid down here.
Cat's speed. There may be two encounters where it will not show any benefit however there are many more encounters where it will be extremely beneficial. People do keep two sets of boots ( I know I do) for the two enchants but that doesn't preclude making sure that you are at least at the hit cap when using the set of boots with runspeed. Since your set is literally at the spell hit cap when using Icewalker, you will be under the cap when using a set with runspeed, thus invalidating your results entirely.
Professions. Don't forget about tailoring and engineering, they also have non-standard benefits that may actually end up being superior to the other options.
Well I already said i didn t know how to handel the config file, so thanks for your advice - I have changed it now.
The race doesn´t seem to chance anything (still getting 7430 dps as Tauren) as long you are not wearing axes, (strange because auf the orc racial), so i forgot to swap it back to tauren,my fault here.
Well I already said i didn t know how to handel the config file, so thanks for your advice.
The race doesn´t seem to chance anything (still getting 7430 dps as Tauren) as long you are not wearing axes, (strange because auf the orc racial), so i forgot to chance it back to tauren,my fault here.
As for the integrity of the Enhsim versus other sims, I will only say that Enhsim is as accurate a sim tool as the community is currently able to maintain. It provides a largely accurate testing environment and given the support provided by many people, bugs are often identified and corrected (such as currently occuring with the matter of swing timers and non-instant casts). Simcraft, according to one of the authors, is providing results with around 1% standard deviation from Enhsim, so I am sure that this is also a valuable tool. Rawr is not a useful tool for us as its since you cannot dictate an ability priority list.
Can you please prove your words that "It provides a largely accurate testing environment"?
I would be very interested in methodology of evaluating simulators accuracy in order to make such claims.
Do you have huge data samples of fully buffed raids beating dummies for 10 minutes?
In fact community is not able now to maintain bug-fixing frequency that Tukez was maintaining and no one knows how many bugs are left.
My point is - this maxdps number and BiS list are completely dependant on current version of one program (if next update will lower the maxdps number - you are going to lower it too, are not you?) and it has to be stated very clearly, not hidden into disclaimer part alongside with minor conventions.
Can you please prove your words that "It provides a largely accurate testing environment"?
I would be very interested in methodology of evaluating simulators accuracy in order to make such claims.
Do you have huge data samples of fully buffed raids beating dummies for 10 minutes?
In fact community is not able now to maintain bug-fixing frequency that Tukez was maintaining and no one knows how many bugs are left.
My point is - this maxdps number and BiS list are completely dependant on current version of one program (if next update will lower the maxdps number - you are going to lower it too, are not you?) and it has to be stated very clearly, not hidden into disclaimer part alongside with minor conventions.
The complaint about not being able to maintain the source is not true. The community has been fixing bugs in the enhsim code since the source has been publicly available all along. Just recently, the armor pen calculations were changed to reflect the Ghostcrawler post on the same subject.
As far as comparing other simulations, it seems to be a worthwhile endeavor but ultimately outside of the scope of this thread. The point of the thread is, as Rouncer noted, about the gear, not the specific number which will never be attained anyway, even on a fully debuffed target dummy. An easier way to check this would be to try and get Rawr or simcraft to somehow disagree with the gear list in the thread. Even if the numbers are slightly different, but the maxdps gearset is the same, it's a non-issue. That's a lot easier than getting a raid together to fight a target dummy.
The complaint about not being able to maintain the source is not true. The community has been fixing bugs in the enhsim code since the source has been publicly available all along. Just recently, the armor pen calculations were changed to reflect the Ghostcrawler post on the same subject.
As far as comparing other simulations, it seems to be a worthwhile endeavor but ultimately outside of the scope of this thread. The point of the thread is, as Rouncer noted, about the gear, not the specific number which will never be attained anyway, even on a fully debuffed target dummy. An easier way to check this would be to try and get Rawr or simcraft to somehow disagree with the gear list in the thread. Even if the numbers are slightly different, but the maxdps gearset is the same, it's a non-issue. That's a lot easier than getting a raid together to fight a target dummy.
1) the purpose of this thread is BiS gear found in Enhsim , that what is used. It is pointless to protect it's "universal" status.
3) Simcrafft team uses BiS gear for reports such as this SampleOutput - simulationcraft - Google Code
but I do not know if they find it with simcraft or export from other sources
4) and as you asked (Off-topic): Downloads - enhsim - Google Code there were 64 Tukez' updates over 9 month (7 per month), comparing to EnhSim - Release: v1.8.1 5 updates + 2 (v 1.7.0.1 and 1.7.0.2) over 3 month (2.3 per month), so my claim about update frequency being much lower now is correct.
Sim's current accuracy is not meassured with the number of past updates, one may claim that current version is absolutly correct ("not guilty until proven" style). Number of updates can only hint to how much work is needed to make it better and better.
Donadon, thank you for your efforts but the combat log you have submitted was invalid and when corrected, netted about 7311 DPS.
Yo! wrote:
My point is - this maxdps number and BiS list are completely dependant on current version of one program (if next update will lower the maxdps number - you are going to lower it too, are not you?) and it has to be stated very clearly, not hidden into disclaimer part alongside with minor conventions.
With regards to the integrity of Enhsim, I go to great lengths to clarify its role in this project. To quote the introductory post:
How is ‘highest potential dps’ defined?
For the purposes of this exercise, ‘highest potential dps’ is defined as the maximum possible DPS achievable within the current version Enhsim DPS simulator created Turkez (EnhSim - Home with the latest executable authored by Sylvand). [...]
It is important to understand the limitations of simulation. Simulation offers a glimpse of character performance under perfect conditions- set latency, no human reaction delay, no system lag, a perfect rotation on a stationary target for thousands of hours. Your actual results will vary greatly from what you observe in the sim, but the sim will give you a clear indication of maximum DPS potential.
I explain that 'highest potential DPS' is defined as that achieved within Enhsim, and I then go on to explain the limitations of simulation (your actual results will vary greatly). I do not think an entire section devoted to this matter constitutes an effort to conceal the nature of our methodology. The information is neither hidden nor along side 'minor conventions'.
Coming back to Cat's, I will take some time to consider some alternatives to the way we are currently approaching things.
3) Simcrafft team uses BiS gear for reports such as this SampleOutput - simulationcraft - Google Code
but I do not know if they find it with simcraft or export from other sources
I think you are comparing a number of things here that are quite different. This thread is not devoted solely to BiS gear but also to a BiS setup that supports that gear (gems, enchants, talents, glyphs, rotations etc). Rawr does not give us a chance to manipulate our ability priority order, so its DPS output is going to be based on a static priority list that may not always suit our gear (PVP totem providing a buff from LL for instance), our Glyphs (ES_SS priority would be affected by SS glyph) and ability priority order (MW3_CL results in huge OOM time if mana talents are not present). Given this, we can at least agree that Rawr is insufficient as a DPS measurement tool and that its results are going to be inaccurate.
I will admit that I am unfamiliar with Simcraft, however one of the Simcraft authors recently posted that his results were within 1% of Enhsim results. We can then generally say that Enhsim and Simcraft are extremely similar in their modelling of the game environment, so we can discuss it in the same way we do Enhsim.
Since Enhsim has a year of updates on an in-game environment that has changed little in that time (the core, underlying mechanics of WoW have not changed greatly) and since the Sim is still being updated atm (recent ArP calculation changes, v1.8.1 changes to swing timer resets on casts [still being tested]) I think we can say that it is a fairly accurate tool that for our purposes.
Obviously the tool is never going to be 100% accurate, but given that no superior simulation tool currently exists, that its functionality meets our needs and that it has survived for a year as a respected, reviewed and updated tool, I think we can say that is the most appropriate and accurate tool for a project such as this. Again, I will say that I make our reliance upon this tool very clear in the introductory post.
If you are able to point us to a more accurate DPS simulation environment/tool I would be keen to explore that as a new standard for this thread.
Raising the minimum mana amount required to trigger Shamanistic Rage will also improve the reported EnhSim DPS. I would cite the exact increase, but I'm using EnhSim 1.8.1 and my output exceeds your reported value with your config. That said, I've always found 750 to be too low when my OOM time was greater than 0.
Try setting it to 2000, this value brings the OOM time to 0.01%. (from your reported 0.18%) The greater the initial OOM %, the more impact increasing this time represents, but don't go past 0% as it will result in a decrease to reported DPS.
As per the instructions I did sim this in the 1.7.0 to calculate mana, however I came across a problem with it whereby I would never get below 2% out of mana time.
I thought this strange and tried it with the current "Max DPS" gear and spec setup and it simmed the same so perhaps I've done something wrong in my config.
Any feedback would be great.
Rawr does not give us a chance to manipulate our ability priority order, so its DPS output is going to be based on a static priority list that may not always suit our gear (PVP totem providing a buff from LL for instance), our Glyphs (ES_SS priority would be affected by SS glyph) and ability priority order (MW3_CL results in huge OOM time if mana talents are not present). Given this, we can at least agree that Rawr is insufficient as a DPS measurement tool and that its results are going to be inaccurate.
As the person who brought Rawr up to patch 3.0+ standards I'd be happy to work with you to improve Rawr to cater for what you believe to be missing. At present it uses the standard priority list as commonly used in EnhSim and defaulted to in my other contribution to the Enh Shaman community ShockAndAwe.
I am perfectly capable of adapting Rawr to cater for what you need. I would however agree that at present EnhSim is likely to give better results as its a SIMULATOR and NOT a closed form solution. Part of my testing of Rawr has involved checking its output vs EnhSim. This has mostly lead to improvements in Rawr but from time to time I find bugs in EnhSim and due to my creation of the project on Codeplex and soliciting help to get someone to maintain it we now have a community maintained version of EnhSim where we can get bugs fixed faster.
Yo! However is right we have no independent verification that the Sim is producing the right results. So the BiS calcs can only be said to be BiS according to EnhSim.
It would also be a good idea to maintain say a top 5 set of "BiS" models. So that whenever there was a new version of EnhSim out you could always re-check the top 5 to see if there were any changes as its entirely possible that the current BiS could be knocked off if a bug is found in the sim.
Originally Posted by Vespasian
Since Enhsim has a year of updates on an in-game environment that has changed little in that time (the core, underlying mechanics of WoW have not changed greatly) and since the Sim is still being updated atm (recent ArP calculation changes, v1.8.1 changes to swing timer resets on casts [still being tested]) I think we can say that it is a fairly accurate tool that for our purposes.
This simply isn't true. EnhSim came into existence during Sunwell Raiding after Yo!'s sim which was the primary sim everyone was using up until that point had ceased to be current. BTW for that reason alone you'd be wise to understand that Yo! does know what he is talking about when it comes to Simulators. Tukez then actively maintained it until March/April this year since when he has been afk.
After much wailing and gnashing of teeth about lack of updates an update came out in the shape of version 1.7.0.3. Then in order to promote a community shared source I created the EnhSim codeplex site and hosted the latest source there. Sylvand then a couple of weeks ago agreed to update the source.
Thus it is wrong to say that EnhSim has had a year of updates, it was heavily maintained up until mid April then a complete lack of updates until mid June then nothing for a month until v1.8.0 and the latest changes this weekend past. ie: only one minor update in 3 months until Sylvand came onboard and did his updates recently.
Last edited by Levva : 07/23/09 at 1:52 PM.
Author of ShockAndAweEnhancement Shaman max dps addon
Please use the EnhSim by Ziff & others to simulate what gear, priorities etc are the best dps. You can use ShockAndAwe to export your paperdoll stats to EnhSim.
I looked at this issue in great detail as the math that was provided on the TTT was totally incomprehensible to me and I was deferring to community consensus on the issue rather than to a clear understanding based on numbers. I decided to go back and do my own. I will be the first to admit that my math is generally pretty unreliable, so please correct it as you see fit. I do anticipate that at least some of the math in this post will be incorrect.
The following post is EXTREMELY simplistic in its explanation, bordering on patronising. Believe me that its style is for my benefit alone, and I apologise for how laboriously it reads.
Run speed provides a speed increase of 8%. Broken down:
If you are running for 1 second, you will get to your destination 0.080 seconds faster.
If you are running for 1 minute, you will get to your destination 4.8 seconds faster.
If you are running for 5 minutes, you will get to your destination 24 seconds faster.
If you are running for 10 minutes, you will get to your destination 48 seconds faster.
Using a BiS set up (roughly 7400 DPS), Cat’s Swiftness over Icewalker nets a 20DPS decrease. Broken down:
Over a minute encounter, this is 1200 total dmg
Over a 5 minute encounter this is 6000 total dmg
Over a 10 minute encounter this is 12000 total dmg
Using a 7400 DPS build, we can then ask how much extra DPS time we need on the target to off-set this lost 12000 dmg?
12000 dmg lost/7400 DPS (how much DMG you do) = 1.6 seconds.
So in a 10 minute encounter:
If you GAIN an extra 1.6 seconds worth of DPS time on the target as a result of your Cat’s Swiftness enchant, then you have not lost any dmg- you have broken even versus Icewalker.
If you do not gain an extra 1.6 seconds on the target, then you have in fact lost dmg vs Icewalker. Icewalker would have given you 12000 dmg, where as your run speed has given you less.
If you gain more than 1.6 seconds on the target, then you have gained more DPS than what Icewalker would have been able to provide. Icewalker would have given you 12000 dmg, but your ability to DPS the boss for longer means that you will do more total dmg.
We must then work out how much running we have to do to for run speed enchant to get us to our destination 1.6 seconds faster. This 1.6 second GAIN must logically be 8% of the total non-DPS travel time (since run speed is an 8% speed increase). So over a 10 minute encounter, that equates to:
X/100*8=1.6 seconds
X= 20 seconds
So:
If during a 10 minute encounter you are geared to do 7400 DPS and you spend 20 seconds travelling between DPS targets, Cat's Swiftness will get you to your target 1.6 seconds faster than if you did not have the enchant, and that 1.6 seconds worth of DPS time would equate to 12000 dmg, the same amount of dmg that Icewalker provides over the course of 10mins.
Travel for more than 20 seconds and you gain even more total dmg on your target (more than the 12000 dmg provided by Icewalker) and so the run speed enchant is superior.
Travel for less than 20 seconds and the run speed enchant has not provided the required 12000 dmg, making Icewalker better.
I wanted to apply these results to Ulduar. The following is a list of Ulduar encounters and the approximate travelling time an Enh Shaman would be forced to endure. All listings are for the hard modes. I am aware that these values are highly subject to both your guild’s strats and RNG (getting chain gravity bombed on XT for instance). They are however entirely accurate in my experience, and I will guess that most people have had similar experiences.
Encounter
Duration
Avg run time
Explanation
Razorscale
5m
0-10s
Running between groups of adds. Running to Razor when she lands..
Ignis
5m
0s
You are constantly on the boss
XT
10m
0-20s
Running out for gravity bombs
Iron Council
8m
10-20s
Running out for 4 (four) overloads/runes of death (depending on strat)
Kolargon
5m
0-10s
Running back to boss after being grabbed
Auriaya
5m
5-10s
Running back after fears
Hodir
3m
5-10s
Running between Frozen targets on the pull, then running to boss.
Thorim
7m
0-10s
Running out to meet an add to begin DPS early (as you walk it back to the middle).
Freya
10m
30s+
Running between various adds and trees
Mimiron
9m
30s+
Avoiding his AoE, chasing adds in P3
General
9m
0s
You are constantly on the boss
Yogg
12m
30s+
Running between adds, tentacles and the brain rooms.
Algalon
8m
10-20s
Avoiding Cosmic Smashes, running in and out of portals for Big Bang.
In your average 5 minute encounter, someone pushing 7400 DPS will need to be running for more than 10 seconds to see run speed as the superior enchant. At a more realistic 6k DPS, the shaman will need to be running for 12.5 secs to reach the break even point. In the case of many examples listed above, the run speed provides either a breakeven benefit or no benefit at all. The only examples where run speed clearly provides superior DPS are Freya, Mimi and Yogg. The rest have very limited run times that are dependent upon luck of the draw, and even then your number may number come up (admittedly, your number might be also be spammed, in which case the run speed would be superior).
Conclusion
I set out to challenge the consensus that run speed will provide superior DPS on most encounters and should therefore be considered the single BiS enchant. Looking at these numbers, I am not convinced that run speed provides a clear, superior DPS benefit over Icewalker in a majority of encounters because:
1) Many encounters derive no benefit from the enchant whatsoever;
2) Many encounters will see the total dmg gained from run speed being roughly equal to that provided by Icewalker.
I would however note the scalability of run speed vs Icewalker. As your DPS increases, the value of run speed over Icewalker increases greatly. We seem to be a point in the content cycle where the duration of encounters and our DPS intersect to make both of these enchants equally viable. This will change in time as average DPS increases. For now however, I am of the opinion that run speed provides no clear benefit over Icewalker to the point where we must make it a mandatory aspect of this exercise. I will continue to allow Icewalker as an enchant, pending the discussion that my flawed math and generic encounter descriptions will generate.