 |
02/06/10, 1:47 PM
|
#151
|
|
Piston Honda
|
As far as I know, yes.
|
|
|
|
|
02/06/10, 3:42 PM
|
#152
|
|
King Hippo
|
Yep, current SC is right.
You can go look at the class mechanics thread discussing it (don't worry, it's a short thread), but the summary is: recount records blocked attacks as "hits" because they are partial hits. It only categorizes whole blocks as "blocks". In all probability, there was never a crit cap change in the first place, but instead, blocks from attacking from the front.
|
|
|
|
|
02/28/10, 4:06 AM
|
#153
|
|
Von Kaiser
|
After noting the removal of critical suppression in Simcraft, I decided to try using [Whispering Fanged Skull] and [Death's Verdict]. Using the Patchwerk combat style I ended up getting:
[Death's Verdict]: 14590
[Herkuml War Token]: 14584
[Whispering Fanged Skull]: 14583
However, using Helter Skelter I ended up getting:
[Herkuml War Token]: 12137
[Death's Verdict]: 12135
[Whispering Fanged Skull]: 12132
I'm not entirely sure how [Herkuml War Token] is calculated because even on Helter Skelter it is displayed under Constant Buffs instead of Dynamic buffs. Even with a 0.5s swing speed it will take more than 6 seconds to ramp up to 20 stacks. Assuming 100% uptime after the initial ramp up time on a 300s fight, that would still only be 98.3% uptime which is enough below 100% to make [Whispering Fanged Skull] better.
All the sims were done using 10,000 iterations, Simcraft 332.0 and the following profile (with just the trinkets swapped):

druid=edgeicator
origin="D:/Users/Redacted/Documents/Character Files\cat.xml"
level=80
race=night_elf
talents=http://www.wowarmory.com/talent-calc.xml?cid=11&tal=0000000000000000000000000000550202032322010053120030313511203503012000000000000000000
glyphs=shred/rip/savage_roar
actions=flask,type=endless_rage
actions+=/food,type=hearty_rhino
actions+=/cat_form
actions+=/auto_attack
actions+=/snapshot_stats
actions+=/faerie_fire_feral,debuff_only=1
actions+=/tigers_fury,energy<=30,berserk=0
actions+=/berserk_cat,energy>=80,energy<=90
actions+=/savage_roar,cp>=1,savage_roar<=1
actions+=/rip,cp>=5,time_to_die>=6
actions+=/savage_roar,time_to_die>=9,cp>=3,if=buff.savage_roar.remains<=8&dot.rip.remains-buff.savage_roar.remains>=-3
actions+=/ferocious_bite,cp>=5,time_to_die<=6
actions+=/ferocious_bite,cp>=4,time_to_die<=1
actions+=/ferocious_bite,cp>=5,rip>=8,savage_roar>=11
actions+=/shred,extend_rip=1,rip<=4,savage_roar>=3
actions+=/mangle_cat,mangle<=1
actions+=/rake,time_to_die>=9
actions+=/shred,if=(buff.combo_points.stack<=4|dot.rip.remains>=0.8)&dot.mangle.remains>=0&dot.rake.remains>=0.4&(energy>=80|buff.omen_of_clarity.up|dot.rip.remains<=2|buff.berserk.up|cooldown.tigers_fury.remains<=3)
actions+=/shred,time_to_die<=9
actions+=/shred,cp<=0,savage_roar<=2
head=sanctified_lasherweave_headguard,stats=564armor_196ap_114crit_106arpen_183agi_183sta,gems=relentless_earthsiege_20arpen_8agi,enchant=50ap_20crit
neck=sindragosas_cruel_claw,stats=120ap_60haste_68arpen_102agi_102sta,gems=20arpen
shoulders=sanctified_lasherweave_shoulderpads,stats=521armor_165ap_90crit_82arpen_136agi_136sta,gems=20arpen_4agi,enchant=40ap_15crit
chest=sanctified_lasherweave_raiment,stats=694armor_212ap_122crit_106arpen_183agi_183sta,gems=20arpen_20hit_6agi,enchant=10str_10agi_10sta_10int_10spi
waist=astrylians_sutured_cinch,stats=391armor_181ap_90crit_74arpen_120agi_136sta,gems=10str_10agi_10sta_10int_10spi_34arpen_20arpen_6agi
legs=sanctified_lasherweave_legguards,stats=608armor_212ap_106exp_122arpen_183agi_183sta,gems=20arpen_10str_10hit_6agi,enchant=75ap_22crit
feet=frostbitten_fur_boots,stats=477armor_181ap_90crit_74arpen_120agi_136sta,gems=34hit_10hit_10exp_6agi,enchant=12crit_12hit
wrists=toskks_maximized_wristguards,stats=304armor_120ap_60crit_68arpen_102agi_102sta,gems=10hit_10str_20haste_4agi,enchant=50ap
hands=aldrianas_gloves_of_secrecy,stats=434armor_165ap_74haste_90arpen_128agi_136sta,gems=10hit_10exp_20arpen_20haste_6agi,enchant=20hit
finger1=band_of_the_bone_colossus,stats=120ap_62hit_65haste_102agi_102sta,gems=20haste_4agi
finger2=frostbrood_sapphire_ring,stats=120ap_60crit_68arpen_102agi_102sta,gems=20arpen_4agi
trinket1=deathbringers_will,stats=167arpen
trinket2=whispering_fanged_skull,stats=148crit,equip=onattackhit_1250ap_35%_15dur_45cd
back=shadowvault_slayers_cloak,stats=185armor_120ap_60crit_68haste_102agi_102sta,gems=34hit_4agi,enchant=23haste
main_hand=oathbinder_charge_of_the_rangergeneral,stats=228ap_114crit_122arpen_195sta_179agi,gems=20haste_20arpen_10hit_10str,enchant=mongoose,weapon=polearm_3.60speed_991min_1487max
ranged=idol_of_the_crying_moon
Last edited by Heef : 03/28/12 at 10:44 PM.
|
|
|
|
|
02/28/10, 8:19 AM
|
#154
|
|
King Hippo
Tauren Druid
Destromath (EU)
|
Originally Posted by Odas
I'm not entirely sure how [Whispering Fanged Skull] is calculated because even on Helter Skelter it is displayed under Constant Buffs instead of Dynamic buffs. Even with a 0.5s swing speed it will take more than 6 seconds to ramp up to 20 stacks. Assuming 100% uptime after the initial ramp up time on a 300s fight, that would still only be 98.3% uptime which is enough below 100% to make [Herkuml War Token] better.
|
Helter Skelter events:
raid_events=casting,cooldown=30,duration=3,first=15
raid_events+=/movement,cooldown=30,duration=5
raid_events+=/stun,cooldown=60,duration=2
raid_events+=/invulnerable,cooldown=120,duration=3
raid_events+=/adds,count=3,cooldown=60,duration=20
raid_events+=/distraction,skill=0.2,cooldown=90,duration=45
Nothing with a long enough duration to let HWT drop
|
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
SimulationCraft Druid Guy
|
|
|
02/28/10, 10:56 AM
|
#155
|
|
King Hippo
|
With three trinkets of different stats with a damage spread of less than 10 DPS, you can assume they're effectively within the margin of error of the tool. I'd definitely say they're all roughly equal.
|
|
|
|
|
02/28/10, 12:16 PM
|
#156
|
|
Von Kaiser
|
Originally Posted by Allev
With three trinkets of different stats with a damage spread of less than 10 DPS, you can assume they're effectively within the margin of error of the tool. I'd definitely say they're all roughly equal.
|
You are probably right, but if theoretical dps almost the same, practical influences of the stats do matter:
War Token: More static Haste, more OOC procs, easier CP generation (but limited by RNG), AP buff almost up all time = less RNG = better, imho easier target switch = better, ramp up time = bad if you have "pauses" (e.g. you cannot attack for some time, like trash, or huge gaps between boss phases), very easy to obtain. Good for Swipe.
compared to:
Death's Choice: More Agi SOMETIMES => easier CP generation SOMETIMES, you have to go PdOk to get it, and its only 10% drop chance, e.g. hard to obtain. Good for Swipe.
Skull: Crit => easier CP generation, easy to obtain (ICC 10 Deathwhisper HM), bad for Swipe.
Did I miss something?
|
|
|
|
|
03/01/10, 10:32 AM
|
#157
|
|
King Hippo
|
Originally Posted by Umah
You are probably right, but if theoretical dps almost the same, practical influences of the stats do matter:
War Token: More static Haste, more OOC procs, easier CP generation (but limited by RNG), AP buff almost up all time = less RNG = better, imho easier target switch = better, ramp up time = bad if you have "pauses" (e.g. you cannot attack for some time, like trash, or huge gaps between boss phases), very easy to obtain. Good for Swipe.
compared to:
Death's Choice: More Agi SOMETIMES => easier CP generation SOMETIMES, you have to go PdOk to get it, and its only 10% drop chance, e.g. hard to obtain. Good for Swipe.
Skull: Crit => easier CP generation, easy to obtain (ICC 10 Deathwhisper HM), bad for Swipe.
Did I miss something?
|
The CP generation should fully be accounted for in the simulator already in the sense that when it's useful to the modeled player, the modeled player takes advantage of it. Also, it's hard to make a case that less individual RNG is better unless you have a specific task you need to carry out that requires a minimum of DPS-- when you have 16 DPSers, the RNG all averages out, for the most part, so occasionally ranking higher on DPS meters might be more rewarding.
How is crit/AP bad for swipe? If any of the trinkets is bad for swipe (without running a sim) it's HWT, because the haste isn't as good as you think it is. You also can't concentrate attacks while a proc is up.
|
|
|
|
|
03/01/10, 8:25 PM
|
#158
|
|
Von Kaiser
|
Originally Posted by Allev
The CP generation should fully be accounted for in the simulator already in the sense that when it's useful to the modeled player, the modeled player takes advantage of it. Also, it's hard to make a case that less individual RNG is better unless you have a specific task you need to carry out that requires a minimum of DPS-- when you have 16 DPSers, the RNG all averages out, for the most part, so occasionally ranking higher on DPS meters might be more rewarding.
How is crit/AP bad for swipe? If any of the trinkets is bad for swipe (without running a sim) it's HWT, because the haste isn't as good as you think it is. You also can't concentrate attacks while a proc is up.
|
The simulator only simulates single target or a very specific setup (multi target of simulationcraft).
Less RNG is always better if the DPS is about the same. It "guarantees" something. If your dps is about the same if you chose crit OR hit, hit is also better (but its lower most times), imho.
Crit/AP are not bad, but Skull is an AP proc. Question is, is static AP (Death'S Choice) or static CRIT better (Skull) is better for Swipe? Token is FAR better because the AP part is up all the time (after ramp up) and the haste part gives nice OOC procs which you require to Swipe a lot, imho.
|
|
|
|
|
03/02/10, 1:39 AM
|
#159
|
|
Von Kaiser
|
I think you're overvaluing the Token. Both versions of the simulation pretty much assumed 100% uptime. In order to even get 99% uptime you'd need something on the order of an 11 minute fight to counter the initial ramp up time not even considering if the buff ever falls off (.5s swing, 1s specials, 20/3s ramp time). Even assuming 99% uptime, which is better than ideal, results in a loss on the order of 20dps, much larger than the calculated difference between the trinkets.
|
|
|
|
|
03/02/10, 4:59 PM
|
#160
|
|
King Hippo
|
Originally Posted by Umah
The simulator only simulates single target or a very specific setup (multi target of simulationcraft).
Less RNG is always better if the DPS is about the same. It "guarantees" something. If your dps is about the same if you chose crit OR hit, hit is also better (but its lower most times), imho.
Crit/AP are not bad, but Skull is an AP proc. Question is, is static AP (Death'S Choice) or static CRIT better (Skull) is better for Swipe? Token is FAR better because the AP part is up all the time (after ramp up) and the haste part gives nice OOC procs which you require to Swipe a lot, imho.
|
First, you can set up a ton of different multi-target simulations if you'd like. You have a lot of flexibility if you know how to edit the text files.
Less RNG is actually only better half of the time. The other half it's worse. Less RNG guarantees a lower maximum at the same time it guarantees a higher minimum. The more RNG-dependent elements you have at the same time, the less it matters. If you have a 25-man raid attacking a single target, RNG will have approximately no effect on raid DPS-- everyone's RNG will "average out" very quickly. This is like rolling 1 die at a time and finding the average. If you roll once, you could get 1 or 6 as much as you'd get a 3 or 4. If you roll 25 times, your average will almost certainly be between 3 and 4 (despite the RNG factor being identical in each roll).
So that's the "RNG averages out faster than you think" part of my discussion. Let's think about it in the context of the carnival game where you swing the hammer and if you hit it hard enough, it rings the bell at the top. Say you got 80% towards the bell, plus or minus a variable due to RNG. If your variable is 20%, you sometimes win the prize. If your variable is 10%, you never ring the bell. Similarly, higher RNG can cause success in a raid when your under the minimum threshold for success, but RNG can sometimes push you over. On the other hand if you swing 110% and your RNG variable is 20%, then sometimes RNG can hold you back. But the majority of fights we see are tuned where player DPS potential is 150% or 200% of what's necessary. The 10% variation wouldn't matter unless you have execution failures, which are as likely to bring you down to 80% as they are to bring you to 120% and suffer from RNG.
You can look back at Vashj-style fights which makes an individual success or failure matter and requires a high minimum DPS, but those fights don't actually exist in WotLK. XT-002 normal mode put some AOE stress on some individual casters, but that was an execution check. I'm trying to think up another minimum-threshold fight, but I can't think of one.
In other words, how often have you looked at DPS to find the cause of a wipe and been able to legitimately say that "DPS RNG was unlucky", instead of "DPS execution failed"? The fights don't require minimum DPS, so at the end of the day it doesn't matter.
Finally, I'd get into how proc mechanics are more often beneficial than build-up procs, but I won't bother. While you're increasing OoC procs, keep in mind that the lion's share of DPS you gain from Haste is white damage. And your white damage is likely to be much lower because all the typical debuffs are less likely to be on your target. I won't argue whether you can count on debuffs being up, other than question an inconsistent philosophy here where you support the RNG of OoC as opposed to more reliable things like crit or an AP/crit proc.
|
|
|
|
|
03/08/10, 5:50 PM
|
#161
|
|
Glass Joe
|
Wrong thread, please delete.
Last edited by minekomineko : 03/08/10 at 9:17 PM.
|
|
|
|
|
03/10/10, 11:00 PM
|
#162
|
|
Glass Joe
Night Elf Druid
Fizzcrank
|
This thread still seemed somewhat active, and being a new Feral Druid I found the information in it quite useful and wanted to work on building an Ovale (and then Faceroller) script to help me learn the rotations.
I see the discussion has moved a little off from the "Best Feral DPS in SimulationCraft" but here's what I came up with anyway.
I got SimCraft up and running and ran the OP's "Best Shred Profile" and made some extremely minor changes (and combined some lines into one).
My settings in SimCraft are:
Patch: 3.3.3
Latency: Low
Iterations: 10000
Length: 300
Fight Style: Patchwerk
Smooth RNG: No
Buffs: All checked
Debuffs: All checked
actions+=/faerie_fire_feral,debuff_only=1
actions+=/tigers_fury,energy<=40,berserk=0
actions+=/berserk_cat,energy>=80,energy<=90
actions+=/savage_roar,cp>=1,savage_roar<=1
actions+=/savage_roar,if=buff.combo_points.stack>=3&buff.savage_roar.remains-dot.rip.remains<=3&buff.savage_roar.remains<=9
actions+=/shred,cp<=4,extend_rip=1,rip<=3
actions+=/rip,cp>=5,time_to_die>=6
actions+=/ferocious_bite,cp>=5,time_to_die<=6
actions+=/ferocious_bite,cp>=5,rip>=8,savage_roar>=11
actions+=/mangle_cat,mangle<=1
actions+=/rake,time_to_die>=9
actions+=/shred,if=(energy>=80|buff.omen_of_clarity.up|buff.berserk.up|cooldown.tigers_fury.remains<=3)
actions+=/shred,time_to_die<=9
actions+=/shred,cp<=0,savage_roar<=2
Running the OP's profile in it's entirety I get:
OP's: 13995
New: 14038
I ran several others in this thread (including my own druid) and would consistently get higher DPS with this action list. I merged the shred lines, raised the SR remains time, raised tigers fury to 40.. and probably a couple other small things. It's not a huge difference in DPS, only 43.. but it was still an increase and so it served as a building platform for my scripts.
I also see the OP instructions aren't updated to the latest patch in SimCraft either so this helps show that the rotation is still solid in the latest updates.
I'm certainly not an expert like others here.. this was just what I got with playing with SimCraft for a few days.
|
|
|
|
|
03/21/10, 4:28 AM
|
#163
|
|
Bald Bull
Orc Warrior
Black Dragonflight
|
Does anyone know how to force the simulator to use mangle as "anytime" ability? I can't get it to use mangle any shorter than the debuff's duration. Yes, I know it's probably not comparable to shred, but I'm just curious for curiosity's sake. The amount of time bosses turn around is getting annoying too.
|
|
|
|
|
03/22/10, 10:33 AM
|
#164
|
|
Glass Joe
Tauren Druid
Boulderfist (EU)
|
Originally Posted by fatalsaint
This thread still seemed somewhat active, and being a new Feral Druid I found the information in it quite useful and wanted to work on building an Ovale (and then Faceroller) script to help me learn the rotations.
|
In the "Visualising the optimal cat rotation ingame" thread there has been some extensive work on writing an Ovale script for feral dps. You can check that thread and see if you find what you are looking for there.
|
|
|
|
|
03/22/10, 11:43 AM
|
#165
|
|
Glass Joe
|
Wrong thread, please delete
Last edited by Propel : 03/22/10 at 12:09 PM.
|
|
|
|
|
|