+- SWTOR Mechanics Forums (http://mmo-mechanics.com/swtor/forums)
+-- Forum: Class Discussion (/forum-7.html)
+--- Forum: Sith Inquisitor and Jedi Consular (/forum-9.html)
+--- Thread: DPS Excel calculator for Sorcerer / Sage (/thread-610.html)
RE: DPS Excel calculator for SI Sorcerer - CaseyTheRetard - 01-21-2012 05:24 AM
Talents!L23 is "2", should be "=IFERROR(P23+O23,0)".
RE: DPS Excel calculator for SI Sorcerer - lostdummy - 01-21-2012 06:38 AM
(01-21-2012 05:24 AM)CaseyTheRetard Wrote: Talents!L23 is "2", should be "=IFERROR(P23+O23,0)".
good catch - it got changed in ver 1.40 it seems
RE: DPS Excel calculator for SI Sorcerer - lostdummy - 01-24-2012 11:58 PM
Posted 1.44 version, most noticeable change is to effect of Wrath and Deathmarks on Crushing Darkness.
Also improved Character tab, simplified buffs selection and made in game stats more visible (since I often compare those between excel and game)
RE: DPS Excel calculator for SI Sorcerer - Caltiom - 01-25-2012 06:06 PM
I found a small difference between the spreadsheet and the sim:
According to http://sithwarrior.com/forums/Thread-Sorcerer-Sage-Mechanics-and-Quirks resource requirements are "rounded up" after modifiers. I first interpreted the wording as a ceil function and implemented it that way in SimulationCraft. Last night I noticed the difference, for example Force Lightning / Telekinetic Throw uses 28 Force in the sim and 27 in your spreadsheet.
So I sent Kor a PM and he confirmed that by "rounding up" he meant ceiling it, although it has been some time since he last tested it.
Another very very small difference arises because of the Crushing Darkness Surge calculation in the spreadsheet. The damage number after surge before any specials is off by about 20 damage. I couldn't figure out why exactly the formula for the averaged surge bonus is dd-surge + 0.6 * extra-td-surge.
Edit: To be a bit more precise: dd * ( 1 + crit_chance * 0.78 ) + td * ( 1 + crit_chance * 1.08 ) is slightly slightly greater than ( dd + td ) * ( 1 + crit_chance * 0.96 )
Edit: They way Effifacy is implemented, am I right that it might happen that there are some Force Lightnings that are casted too much when hitting the point of being below 27/28 Force, which wouldn't allow another cast even if you'd get part of your Force back later?
RE: DPS Excel calculator for SI Sorcerer - lostdummy - 01-25-2012 07:15 PM
(01-25-2012 06:06 PM)Caltiom Wrote: I found a small difference between the spreadsheet and the sim:Thanks, will change this to ceiling - and also will try to check ingame if you are not certain about this. Problem with ingame tooltip is that we actually do not know if game only "shows" rounded value, or also "use" rounded values ;p
According to http://sithwarrior.com/forums/Thread-Sorcerer-Sage-Mechanics-and-Quirks resource requirements are "rounded up" after modifiers. I first interpreted the wording as a ceil function and implemented it that way in SimulationCraft. Last night I noticed the difference, for example Force Lightning / Telekinetic Throw uses 28 Force in the sim and 27 in your spreadsheet.
So I sent Kor a PM and he confirmed that by "rounding up" he meant ceiling it, although it has been some time since he last tested it.
(01-25-2012 06:06 PM)Caltiom Wrote: Another very very small difference arises because of the Crushing Darkness Surge calculation in the spreadsheet. The damage number after surge before any specials is off by about 20 damage. I couldn't figure out why exactly the formula for the averaged surge bonus is dd-surge + 0.6 * extra-td-surge.Before version 1.44, excel used fixed ratio of DoT/DD for Crushing Darkness, but in 1.44 it was changed to use actual (ie to account for LingeringNight).
Edit: To be a bit more precise: dd * ( 1 + crit_chance * 0.78 ) + td * ( 1 + crit_chance * 1.08 ) is slightly slightly greater than ( dd + td ) * ( 1 + crit_chance * 0.96 )
With this change to DoT part in 1.44, I see 98% as combined surge value, instead of 96% (although real values are not rounded) - so if you checked this before 1.44, that would explain difference.
(01-25-2012 06:06 PM)Caltiom Wrote: Edit: They way Effifacy is implemented, am I right that it might happen that there are some Force Lightnings that are casted too much when hitting the point of being below 27/28 Force, which wouldn't allow another cast even if you'd get part of your Force back later?That very much depend on order in which spells would be used, so I presume it could happen but very rarely. It wont ever happen for rotations that end fight with above 25 force (which is true in most cases, and for all builds currently). And even for rotations/builds where fight ends with between 0 and 25 force, if last used spell in fight would be any using 20+ force (practically any but FL) , that edge case would not happen. And excel does not define exact order in which spells are used, only how many times.
RE: DPS Excel calculator for Sorcerer / Sage - Kaedis - 01-26-2012 01:29 AM
Quote:Thanks, will change this to ceiling - and also will try to check ingame if you are not certain about this. Problem with ingame tooltip is that we actually do not know if game only "shows" rounded value, or also "use" rounded values ;p
Hmm...this could be an issue. However, if the in-game values were still set to non-integer values and used as such in the calculations, I would expect the tooltips to either use standard rounding or flat truncation. Ceiling would be rather abnormal for that setup.
RE: DPS Excel calculator for Sorcerer / Sage - Armageddus - 01-26-2012 10:37 AM
Hi guys, first post here, but I've lurked quite a bit. We've been clearing our way through all the hardmodes and will probably start fiddling with Nightmare next week. All of my experience is PvE-based.
I apologize if I missed this, but I read through the thread and I can't find where Conduction was derived to produce 0 DPS gain (the workbook doesn't factor a gain when I drop the point in it). I did see some discussion on uptime on it, but missed it if it was in there. I have found from the field that my uptime on it is quite high, not 100 %, but I rarely never see at least one stack of the buff.
I initially started out with a lightning-heavy build and found fights required too much movement to really hard cast as much as that build desires to max out its DPS. I then switched to the hybrid 13/28 build and it did feel noticably different in damage output, but I ended up starved for resources during long fights as we progressed through hardmodes.
I went exploring on what I could do to get a build whose damage is consistently high regardless of whether I can sit still and nuke or am running around all over the place, and where I never run out of resources. I haven't seen anything like this, and the workbook doesn't compare it as favorably as the other builds, but I'm running a 25 lightning, 16 madness build that works around instant-cast abilities (primarily chain lightning), lightning barrage, and lightning strike. It has worked very well for me and even though I'm not getting the BIG NUMBERS!, I definitely feel over the course of boss fights my damage is high, stable, and sustainable.
Madness tree has Calcify, Will, Disintegration, Chain, Madness, Efficacy, Death Field (DPS loss according to sheet?), Madness, and Wrath.
Lightning tree has Convection, Reserves, Induction (only 2 points), Barrier, Subversion, Exsanguinate, Spire, Barrage, Chain, Storm, Effusion, Forked, Polarity, and Conduction.
So far I've been loving it... I start off stacking my buffs off strike casts with shift, then it becomes management of the instant-cast procs (one is always up between wrath and storm, many times both are up). Barrage procs quite a bit and is basically the primary hard-cast, with strikes in between while waiting for another barrage (also keeps stacks going). If I have to move, instant chains, instant strikes, instant crushes, and death field if nothing is up do the trick. Love death field because I can cast it behind me while running lol.
The spreadsheet doesn't make it look as good as current standards, but I noticed some discussion of the implications of fight mechanics and movement necessary - and I can attest from ground experience that is what drove me to spec this way and it is working wonders.
I lack the time to crunch the hard-corp numbers on this spec, but if someone wants to give it a whirl in some of these encounters maybe they can prove my gut-check that this build is very strong no matter what.
Besides that thank you all for the hard work you put into this. Keep rockin' on!
RE: DPS Excel calculator for Sorcerer / Sage - lostdummy - 01-26-2012 07:47 PM
Interesting build - and as long as builds are fairly close to each other in DPS, it is best to use one that suits your gameplay. I also do not use 13/28, but slightly modified version with Lightning Effusion.
Related to some of your comments, I set build you mentioned as 'custom', and :
(01-26-2012 10:37 AM)Armageddus Wrote: I can't find where Conduction was derived to produce 0 DPS gain (the workbook doesn't factor a gain when I drop the point in it). I did see some discussion on uptime on it, but missed it if it was in there. I have found from the field that my uptime on it is quite high, not 100 %, but I rarely never see at least one stack of the buff.Conduction gives no DPS gain on this build since excel use Force Lightning as preferred filler (better DPCT than LS) - and if LS is not used, no conduction.
You can easily set excel to follow your rotation - since you said you use FL only under barrage, and LS otherwise, just set '0' in USE field for FL (S16). That would result in slightly lower DPS (~5%), as well as more force limited case - meaning if it would be simple tank&spank 5min fight, you would have to reduce use on CLs to keep up force if you use LS instead of FL. Because of that, excel also may remove CL from rotation, but you can also prevent that if you set '0' in 'Limit CL' on Options tab. Alternative is to set '0' to Auto-rotation on Options tab, and then leave '1' in USE field only for those spells you would use in rotation (ie put '0' for Shock too etc).
Anyway, when you set your own rotation instead of one excel auto-suggested, you will end up with quite a lot Conductions - average stack depth would be 2.65, and gain from Conduction talent in such case is 2.65%.
(01-26-2012 10:37 AM)Armageddus Wrote: If I have to move, instant chains, instant strikes, instant crushes, and death field if nothing is up do the trick.I presume you can use mostly just one of those, depending on which effect is active (if any) between Wrath and Storm. When I check excel, it shows 16 Storms and 24 Wraths when LS is used, or 54 Wraths and 0 Storms when FL is used as filler. That would indicate that using FL as filler would even give more chance for insta spell being available when you need to run.
There is one case when having smaller number of Wrath+Storms may be better than having larger number of Wraths - and that depend on what happens when both Wrath and Storm are up. Since I can not test above, I'm interested if you noticed during fights answer to :
When both Wrath and Storm are up, and you cast CL - will only Storm be used, Wrath, or both?
If only Storm is used, then in those rare cases when both of them are up you could use 2 insta spells on run. Depending on how often that happens (that both are up just at moment you need to run), it could make 40 Wrath+Storm better than 54 Wraths. Otherwise, this build would probably have no significant advantage in 'running' type of fights compared to Hybrid build, which also can use all above listed spells, but under (more) Wraths.
And there is another case when Storm+Wrath may be better than more Wraths, and that in cases when those 'running' cases are very short (under 6sec) and very often - in those cases it is possible that if CL was used recently it would not be up even if Wrath is up (and Crushing mostly is not ready). In such cases Storm will reset cooldown on CL , making it more useful than Wrath. But I believe such cases are rare - any running breaks longer than 6sec (and in fact, longer than 3sec as that would be average time to wait on CL cooldown), would still give you chance to cast CL under Wrath. Also, its always possible to cast LS under Wrath if CL is not ready.
(01-26-2012 10:37 AM)Armageddus Wrote: The spreadsheet doesn't make it look as good as current standards, but I noticed some discussion of the implications of fight mechanics and movement necessary - and I can attest from ground experience that is what drove me to spec this way and it is working wonders.Yes, there is option in spreadsheet to set some DPS breaks, and partially simulate fight mechanics where you do not DPS all the time, but so far only in cases where you need to 'hide' in those periods - which is not same as when you need to run.
But as I mentioned before, if build has similar DPS to others in 'ideal' fights, and as long as it better suit someones playstyle, its good one to use.
As to question if this build is better than Hybrid build for 'running' type of fights, I guess answer to above two questions is needed:
- what is used for CL if both Wrath and Storm are up?
- how often are both of them up when running phases occurs?
RE: DPS Excel calculator for Sorcerer / Sage - lostdummy - 01-26-2012 08:43 PM
I tried to see what would be needed to model effects of 'running' , based on above posts and also on previous 'DPS breaks' that I already put in excel for 'hiding'.
While modeling such running breaks is probably best task for simulation, rather than excel, I still wanted to see if I can get at least some indication of what build would have 'advantage' in such cases.
For example, with Hybrid build, on running break I can:
- Shock
- CD, if Wrath up and off cooldown
- CL, if Wrath up and off cooldown
- DF
- Affliction, if close to expire
- LS, if Wrath up and none of CD/CL off cooldown
And since Shock has 6sec CD, that probably means some combination like: Shock, Wrath(CD or CL or LS), DF, Shock, Affliction
Now, different build like above hybrid+storm has same options, with change that "CL, if STORM or Wrath up and off cooldown". Madness build would add 'Creeping Terror' as possible spell to use if off cooldown, and Lightning build would have Thundering Blast as option for Wrath too, if off cooldown.
Another difference between builds is in damage of these spells - Shock under Hybrid has better damage than under Lightning, while LS is reverse.
So basically, I see different builds differ in following for running breaks:
1) available spells to use while running or under Wrath
2) chance that Wrath and/or Storm will be available when break occurs
3) DPE (damage) of available spells
Probably best thing to compare builds for 'running' would be answer to : "what is average damage that can be done under this build on one running break"
For #1 (available spells), some approximation of spell on cooldowns to be ready during break is : p= Break duration (Tb) / Spell repeat period , so if "Ns" is number of spells used, and "Fd" is fight duration (300sec), then:
p_noCD= Tb*Ns/Fd
For #2 (chance for some effect like Wrath or Storm to be available), if "Nse" is number of spells using that effect in regular rotation, and "Neff" is number of triggered effects, some approximate formula is:
p_effectUp= 1- Nse/Neff
And #2 formula, applied to excel numbers for previous Hybrid+storm build, indicate that for example 'storm' would rarely be available for any break, since Nse (number of CLs under storm) is about same to Neff (number of storm effects). That seemed strange to me, but on second thought there is explanation: in normal fight, whenever storm effect appear, you would almost immediately use CL and use that storm (if you want to maximize DPS) - so there would be practically no chance that some storm will be available in break (well, small chance if break happens in that 1-2sec between storm appearing and CL being casted, which is under 5% chance)
For wraths, Nse (number of Crushing or CLs under Wrath) is lower than Neff (total numebr of wrath procs), so almost 50% chance (for this particular case) exists that wrath will be up when break occur - making #1 (chance that spels are off cooldown) limiting factor.
RE: DPS Excel calculator for Sorcerer / Sage - haldirgoth - 01-26-2012 11:47 PM
in the new version 144, it seems with the same stats , lightning is better at dps than hybrid , was that after some recalculations ?