-ammo isn't loaded/added to the config file (add a default value?)
Armory doesn't tell you what ammo is equipped, unfortunately. Good catch though that we're using "0" as a default. I'll switch that to use the equivalent of Saronite Razorheads by default.
Originally Posted by Hagen
there is no pet family set (add wolf as default?)
It is set. It's just set in the action list. When it does "summon_pet" there, it'll summon a wolf unless you use a summon_pet option to override that.
Originally Posted by Hagen
the armor penetration on my Boots of the Renewed Flight wasn't read correctly (was missing)
Looks like the parser was looking for "armor penetration" instead of "armor penetration rating". I'll fix that.
Thanks for the feedback. I notice that you figured out that to do a realm with a space in the name, you need to use url-escaping on the space (ie simcraft armory=eu,Der%20Mithrilorden,Lunaera or simcraft armory=eu,Der+Mithrilorden,Lunaera )
Armory doesn't tell you what ammo is equipped, unfortunately. Good catch though that we're using "0" as a default. I'll switch that to use the equivalent of Saronite Razorheads by default.
It's not shown on the display page, but the XML page source has arrows included as an item with slot="-1".
Interesting. Apparently the US armory shows pet specs too while the EU armory doesn't. That's something we should take advantage of in the future, but for now I changed the armory parsing code to not blow up because of it. If you try 2772 or later, it should work now.
The url for the talent spec isn't correct. Mine was talent-calc.xml?c=hunter, but on the WoW site it's talent-calc.xml?cid=3
For those working SVN and living on the ragged edge, this has been fixed in r2775. For everybody else, I should have a new release out Sunday when I return from vacation.
For those of you who don't mind building from trunk (living on the ragged edge as dedmon puts it), I've added armory pet parsing and it could use some testing. We still only support cats, wolves, raptors, devilsaurs and wind serpents, but if you've got one of those, you can try it with arguments on the form
Earlier this month I did some testing on CSS and found out that SimCraft had it doubledip from +%dmg talents/buffs, so I fixed that, but in the same move made it not benefit from anything: No +crit. (r2767)
I fixed it today after someone filed a ticket on the issue, sorry for that.
Fixed in r2860, now CS and CSS should have the same crit%, as expected.
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
Testing out the armory import, I see that race apparently isn't noted. This matters for racial bonus (e.g. dwarven +1% crit with guns), which would be kinda nice to have for a quick and easy weapon comparison.
Updated the hunter profiles with epic gems etc, done with latest revision (r3141)
Hunter_T8_07_57_07_HM: imp.HM+glyph bitch, to show the kind of dps loss commited to the buff
Thanks for your work. I just have two advices regarding the Hunter's Mark spec:
1. It should be better to change chimera shot glyph for hunter's mark (if you really want the glyph at all).
2. It' better to take points out of Improved Steady Shot and maybe one from Rapid Recuperation than to take them from Improved Barrage to skill Improved Hunter's Mark.
I hate to bump this thread after so little recent interaction with it; however, I've come to a slight dilemma.
The dilemma being that with 3 different ways to compare gear/dps, it's hard to gauge which is the best. Simulationcraft has SV coming out on top while the spreadsheet/web app have MM coming out on top.
In addition to this, different sets of gear bring out different comparisons (i.e. spreadsheet has helm #1 > helm #2 while simulationcraft has helm #2 > helm #1).
I guess the best way to test is to do manual tests on a training dummy and go with real world results rather than tools. But, when building a gear set for the future, one doesn't have access to the gear to test it themselves =/.
I'm inclined to believe that if both tools show believable results but disagree, then the difference is smaller than what you'll be able to notice in practice.
Any number of differences in modelling (e.g. lag) will cause discrepancies. Of course, it would be interesting to hunt down modelling differences. Personally I don't have access to Excel, making that rather hard.
I think there are some inaccuracies in the MM model. I've been hoping to spend some time on it and get it in-sync with 3.2.2 before 3.3 comes out. Among other things, it doesn't seem to have weapon damage calculated correctly, which impacts a number of other calculations (which have a bigger impact on MM than SV). MM definitely is low in simulationcraft right now.
I fixed a bug that caused chimera-serpent to get the partial resists reduction twice, but after I fixed that I got very similar numbers in the sim and in-game. (Very anecdotal, I only tried one gear set, but I did it without proc trinkets and with a low level gun with low damage range.)
As I said, I don't have access to Excel, but I compared to the numbers on femaledwarf.com and noticed that simc's chimera damage is still lower. I have no idea if that means simc's damage is too low or if femaledwarf's is too high. I also don't know how femaledwarf.com treats partial resists.
The scaling of weapon damage is pretty similar between femaledwarf and Simulationcraft (around 8DPS per 1 weapon DPS for me) but the haste scaling is still way off in Simulationcraft (1.4DPS in Simulationcraft vs. 0.7 DPS at femaledwarf).
Has anyone seen similar results or clues to the cause of this?
The scaling of weapon damage is pretty similar between femaledwarf and Simulationcraft (around 8DPS per 1 weapon DPS for me) but the haste scaling is still way off in Simulationcraft (1.4DPS in Simulationcraft vs. 0.7 DPS at femaledwarf).
Has anyone seen similar results or clues to the cause of this?
Using femaledwarf, you can play around with your time-to-wait (for CD on a higher damage shot instead of steady) and latency to get widely varying values for haste. I believe default is 0.8 @ 150ms. I've seen the odd cases where femaledwarf even suggests haste over agility depending on your settings, at which point I just readjust my time-to-wait.
Reproducing a bit of PM conversation we've been having, since Deadmon pointed it out it would be useful to have in public:
Originally Posted by dedmonwakeen
Originally Posted by alienangel
Hi,
I've been trying to tweak my simcraft profile to do the following:
- instead of always firing chimera shot and aimed shot back to back, interject some steadyshots between them. Since a normal unconditional priority sequence of chimera->aimed->steady usually results in 4 steadies between each pair of aimed/chimera, the initial thought was to interject 2 steadies between each chimera/aimed pair.
(the ideal behind this is to maximize utilization of Improved Steady Shot procs from the steadyshots, instead of overwriting them by only firing a chimera (or aimed) after 4 steadies in a row).
I've so far been unable to make a conditional action that achieves this behaviour. I can't find any documentation anywhere on what conditions can be used (and how) on the "actions+=/spell" lines. I've tried poking around various source files on the webSVN, but no luck there either. If a good place to look exists (I am reasonably capable of deciphering code, although my C is a bit rusty) it'd be great if you could point me to it, if not, answers to the following questions would be much appreciated:
- is it possible to count in an action conditional whether a particular spell has just been cast (e.g.: spells_n_minus_1=steady_shot, spells_n_minus_2=steady_shot)?
- is it possible to check in an action conditional for hunters what the cooldown on a particular spell is (e.g.: chimera_shot>=7)?
- is it possible to check in an action conditional for hunters what the duration on a non-dot debuff is (e.g.: debuffs.aimed_shot.remaining>=7)?
If one of the above 3 conditionals (ideally the first) can't be made, I think the only way to model what I want would be modifying/adding ready() functions for aimed, chimera and steadyshot to make them want to fire and not want to fire at particular times :/ I hope this isn't the case, since a counter conditional (or even a straight castsequence to use when moving=0) would be much less messy.
Thanks!
I'm afraid the docs are sparse.... and the action conditionals are rather new.
However..... We do not model the aimed shot debuff at the moment. And we do not have a "previous spell" function available for conditionals either.
Feel free to ask these questions in the SimulationCraft for Hunters thread. I'd love to breathe some life into it and hopefully improve the Hunter module for you guys.
Now, I did try the quoted conditional for chimera shot, by entering the following actions list:
I would think that when it hits the first steady shot action, it would fire it for the first 2 gcds after using a chimera, then stop using it since chimera's cooldown would then be less than 7s, at which point it would do an aimed instead, then more steadies until it can fire chimera again. However what actually seems to happen is that it never stops firing steadyshots, and 0 aimed shots get fired :/
I also tried taking out the first steady_shot action and just putting a condition on aimed shot for chimera_shot's cooldown being <7, but the parser likes that even less (likely something about the syntax I'm failing to understand).
edit: In fact, the parser just seems to refuse conditions on aimed_shot: I can stick a "cooldown.chimera_shot.remaining>=7" on a steady_shot action without any complaints (although the sim result isn't what I expect), but if I try the same cooldown condition on an aimed_shot line, I get parse errors such as:
I should have looked at the code instead of responding from memory.
Key differences:
The "option" to the action is "if" which takes an expression, ie: if=((A<B)|(C>D)|!E)
The notation for dot/buff/debuff/cooldown time remaining is ".remains" and not ".remaining"
I'd tried it with the =if before, but was still getting the parse error, since the "remains" bit was still wrong.
I still couldn't get a condition on the extra steadyshot line to do what I wanted (adding a "actions+=steady_shot,if=cooldown.chimera_shot.remains>=7" after the chimera_shot action somehow results in no steadies, aims, OR chimeras firing at all, very strange) however putting a condition on aimed shot seems to do what I want:
Results in a loss of 15 DPS over the normal sequence, going from 10131dps with no condition to 10115dps with the conditional aimed shot. I see aimed shot's Damage Per Execute going up in the latter run, presumably because of the extra IIS procs it's getting, and Chimera's DPE going down, however the average number of aimed shots fired is going down slightly as well, and the number of steady shots fired is going up slightly - presumably because on the first time through the sequence you delay the first aimed shot.
So if this is right, although the splitting into a 2xSteady-1xSpecial system does seem to increase the average damage of the specials, it's not worth it (at least in my gear) unless you have a much longer fight, or can work into the offset cooldowns without losing any shots you'd have lost anyway.
I think there's also a debug option to get SimCraft to print out the list of actions it's performing, going to try to enable that to confirm it's actually doing the shots as I think it is.