Page 2 of 2 FirstFirst
1
2
  1. #21
    Fluffy Kitten xtramuscle's Avatar
    10+ Year Old Account
    Join Date
    Jun 2010
    Location
    Suffolk, UK.
    Posts
    1,749
    Quote Originally Posted by Teryx View Post
    I finally wrapped my head around how to simcraft openers; it's actually pretty straightforward, all you have to do is change "actions.precombat+=/starfire" to "actions.precombat+=/wrath", "actions.precombat+=/starsurge" etc in the rotation script and compare the numbers on a short fight. The numbers you get is DPS over the whole fight so you should multiply the differentials with the length of the fight in seconds to get the actual damage output difference.

    For me on single target, at 703 with no Archie trinket, ring or set bonuses, the Starsurge opener gives me an average loss of around 250k damage to using Starfire, which is only slightly ahead of Wrath.
    I get that by pre casting starsurge before combat it isn't buffed by CA, but I dont see how with 703 ilv, casting one extra (also unbuffed) starfire + a (buffed) ss during CA nets THAT much of a dps difference. It's unlikely I'd see that gain with my gear 720+ don't see how it's possible with the gear you mentioned.

    tl:dr Would be interested in seeing the log atleast.
    Vexxd

    LFG to push 15+ m+,
    maybe streaming @ http://www.twitch.tv/vexxee

  2. #22
    Quote Originally Posted by xtramuscle View Post
    I dont see how with 703 ilv, casting one extra (also unbuffed) starfire + a (buffed) ss during CA nets THAT much of a dps difference.
    You are right, it's alot. I've given it some thought and experimented further. An important difference to SF vs SS at pull is that SF does not make the eclipse cycle start until the cast is finished, whereas with an SS precast you are one GCD into Lunar at the zero point - I believe that might be the factor behind the simmed differences.

    It's possible that simming short fights becomes so sensitive to changes in the eclipse rythm that it might skew the result, so I've run some 600 second sims with +/-50% fight length. It's all on Patchwerk, 250k parses. Using your profile Vexxe (noone cares about ilvl703 anyways), with the Archie trinket, I get 75980 dps with Starsurge opener and 76781 with Starfire.. In 600 seconds that's a difference in favor of Starfire of 481k damage. Replacing the Archie trinket with Chipped Soul Prism gives 77553 dps for Starfire opener and 76448 with Starsurge opener, which gives Starfire an edge of 663k damage. This is with 10 minute fights, simming only 30 seconds fixed lenght fights gives SF 363K oves SS with Archie trinket and 467K in SF's favor without.

    These experiments should be very easy to reproduce and I really doubt they are ilvl or gear dependant beyond the obvious influence of the Archie trinket. I am, however, not entirely sure on what extent simc can be used to evaluate different rotation rules reliably. Maybe these numbers are a consequence of the algorithm in the default Moonkin script, which doesn't really represent how most people play. If not, time spent in Lunar in the opener might be a thing that actually matters.

    Edit: The difference is so large that real life testing should be able to make it clear without pushing the sample size very high whether these simc predictions are valid or not.
    Last edited by Hildrande; 2015-09-06 at 05:14 AM.

  3. #23
    Fluffy Kitten xtramuscle's Avatar
    10+ Year Old Account
    Join Date
    Jun 2010
    Location
    Suffolk, UK.
    Posts
    1,749
    I'm just not getting those numbers. For example, how do you leap from 'dps numbers' to the 'damage numbers'?

    'simming only 30 seconds fixed length fights gives SF 363K over SS with Archie trinket and 467K in SF's favor without.' is that dps or damage done?
    Vexxd

    LFG to push 15+ m+,
    maybe streaming @ http://www.twitch.tv/vexxee

  4. #24
    Quote Originally Posted by xtramuscle View Post
    how do you leap from 'dps numbers' to the 'damage numbers'?
    By multiplying the differense in DPS by the length of the fight. For the second question, it's damage, not DPS.


    The results are so counterintuitive I'm sceptical.

  5. #25
    Fluffy Kitten xtramuscle's Avatar
    10+ Year Old Account
    Join Date
    Jun 2010
    Location
    Suffolk, UK.
    Posts
    1,749
    Quote Originally Posted by Teryx View Post
    By multiplying the differense in DPS by the length of the fight. For the second question, it's damage, not DPS.


    The results are so counterintuitive I'm sceptical.
    The theory of an opener, is exactly that, it's the sequence of spells on the pull before a normal rotation resumes. If you're multiplying the result by the length of the fight, you're altering our basic rotation, which will not be a dps increase. Essentially you're saying your starfire opener does maybe 100k damage more in the 30 second opener, but because its a 10 minute fight its an increase of (200k a minute x 10) 2mil damage. Which is clearly completely incorrect. (for example).

    If you're saying in a 30 second pull window, pre casting SF instead of SS yields 363k more damage, that may be correct, but personally that value still seems too high to be viable without some weird sort of rng starfire crit.
    Vexxd

    LFG to push 15+ m+,
    maybe streaming @ http://www.twitch.tv/vexxee

  6. #26
    Don't ever CA before your pre-cast starsurge IMO. If you use it before for your starsurge you just lose an extra starsurge at the end under CA, however that extra starsurge at the end will also be under trinkets and ring and contributing to your explosion. It is simple multipliers to choose to have that starsurge when as many buffs are up at once.

    My opener is:

    -4.5: Incarnation
    -3: Pot and Wrath (Macro)
    -1.5: Starsurge
    0: CA and Moonfire (Macro)
    +1.5: Berserking and Starfire (Macro)
    Starfire
    Starsurge
    -
    -
    -

    You know the rest...

    With my haste and crit levels almost certainly will I get enough procs to have a starsurge firing as CA is 0.5 secs from ending. If I had used CA for my first Starsurge instead I wouldn't have got this Starsurge that is under the effects of 1.6K INT from trinket and +25% damage from the ring.

    Not actually fully accessed this opener but I liked to see what every else is deciding is best

  7. #27
    Quote Originally Posted by kayin60 View Post
    I got this opener for Nagura (Method's Druid)

    4.5-5 sec pot/inc -> 3 sec cast wrath and SS right after -> CA/berkserk if your a troll and begin your normal rotation.
    Using berserk before the first SF is a huge waste because it puts you below the gcd cap.

    Quote Originally Posted by Albimoo View Post
    Don't think it would be worth losing potentially an extra buffed starsurge or starfire during CA to send off one starsurge with CA with no trinket procs etc.

    Plus opening like that just wastes time with because the dots tick at least one less time with CA buff.

    Wouldn't be a HUGE DPS loss, but definitely think it would be a loss.
    If you are doing an opener with SS, you might aswell add a wrath in front of it, proccing all the trinkets for you, thus you are only losing the legendary ring, which is why it was the best opener in BRF. (obviously talking about 4.5 inc 3 PP/wrath 1.5 CA/SS.) Seems like the only alteration is to save CA for 0 sec instead for HFC.

    edit: after reading what you replied to he is so far off a decent opener it is sad.

  8. #28
    Quote Originally Posted by xtramuscle View Post
    If you're saying in a 30 second pull window, pre casting SF instead of SS yields 363k more damage, that may be correct, but personally that value still seems too high to be viable without some weird sort of rng starfire crit.
    I've tried look more closely into it and with the SS opener, simc will execute CA after Moonfire, whereas the default Starfire opener script will send off an SS before hitting CA+Moonfire.

    Good stuff eh :-P

    With fixed scripts the results are as expected, SS is slightly ahead of SF, which is even slighter ahead of just Wrath. SS's lead over SF is in the vincinity of 35k with soul prism, 15k without. Simmed, with 716 gear from Michler-Kazzak-EU.

    Wrath+SS can't be simmed straightforward, since simcrafting's clock starts ticking the moment the first spell cast is finished. Raw sim puts Wrath+SS around 40k behind just SS, but by adding 1 second of DPS value (73k on the 100 second sims I ran), Wrath+SS gets a net gain over SS of 33k damage (not DPS).

    For reference, this is the single target command list that gave proper pull execution:
    Code:
    actions.single_target=celestial_alignment,if=eclipse_energy>=0
    actions.single_target+=/moonfire,if=buff.celestial_alignment.up&remains<1
    actions.single_target+=/incarnation,if=eclipse_energy>0
    actions.single_target+=/starsurge,if=buff.lunar_empowerment.down&(eclipse_energy>20|buff.celestial_alignment.up)
    actions.single_target+=/starsurge,if=buff.solar_empowerment.down&eclipse_energy<-40
    actions.single_target+=/starsurge,if=(charges=2&recharge_time<6)|charges=3
    actions.single_target+=/sunfire,if=remains<7|(buff.solar_peak.up&buff.solar_peak.remains<action.wrath.cast_time&!talent.balance_of_power.enabled)
    actions.single_target+=/stellar_flare,if=remains<7
    actions.single_target+=/moonfire,if=!talent.balance_of_power.enabled&(buff.lunar_peak.up&buff.lunar_peak.remains<action.starfire.cast_time&remains<eclipse_change+20|remains<4|(buff.celestial_alignment.up&buff.celestial_alignment.remains<=2&remains<eclipse_change+20))
    actions.single_target+=/moonfire,if=talent.balance_of_power.enabled&(remains<4|(buff.celestial_alignment.up&buff.celestial_alignment.remains<=2&remains<eclipse_change+20))
    actions.single_target+=/wrath,if=(eclipse_energy<0&eclipse_change>cast_time)|(eclipse_energy>0&cast_time>eclipse_change)
    actions.single_target+=/starfire
    Last edited by Hildrande; 2015-09-09 at 04:49 AM.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •