EnhSim, DPS simulator - Page 109

Rouncer · 11/09/09, 9:19 AM

Leave SR as it is currently implemented. It's easy enough to set the mana threshold to say 15000 so that SR becomes top priority to be used whenever it is off cooldown. That also gives us options for how to deal with the bonus and keeps validity for mana figures when we start adding things like the Fire Nova Glyph and Improved Fire Nova Talents.

Popping Fire Nova every 3 seconds is going to use an incredible amount of mana and without control over SR there would be no way to know if it would be at all sustainable. Perhaps it could be used as massive AoE for specific situations by popping SR when you get low on mana. Currently we can manipulate the combat length and mana level setting to determine when it would be best to pop SR for maximum duration without going oom but if you link the bonus to just using SR as a dps cooldown that option would be lost.

As for the Fire Elemental, it has 3 attacks. An AoE nova, a damage aura and a direct damage attack. Beyond that we will need testing to determine how he scales with spellpower (he does scale) and other stats as well how those abilities are used. Strawberry said he saw a 5 second cooldown on the Nova but we could use a lot more specific testing. We also need that testing to be on the Boss Dummy as well as little ones so we can see the Elemental's miss rate and determine if that is affected by our stats.

Fire Nova is exactly the same as the current totem except it becomes an activated instant ability with a 10 second cooldown. It should have a 21.42% coefficient. Questions are if it will still be affected by Call of Flame and whether it can proc and/or be affected by Elemental Focus (there were early reports that it could proc it but was unaffected by it). Fire Nova does work with the Fire Elemental Totem with the Nova originating off of the totem not the Elemental.

Zimeron · 11/09/09, 9:54 AM

A couple people tested out the Fire Elemental's scaling in the SimC group. Here's what we came up with:

Swing: 180 + 0.6 * SP + ( 2.0 / 14 ) * AP

Blast: ( 725 to 825 ) + 0.2 * SP

Nova: ( 925-1050 ) + 0.5 * SP

Shield: 96 + 0.015 * SP

Base Mana: 4000 + 15(0.3 * SP)

Nova is a two second cast and doesn't appear to have a cooldown.

His AI is also wonky and doesn't always appear to use every ability to it's fullest. For example he'll have .5 second pauses of nothing.

dedmonwakeen · 11/09/09, 10:59 AM

Some notes on my examination of (too many) Fire Elemental parses:

(1) I would agree with Strawberry on the 5sec CD for Fire Nova
(2) While Fire Nova may be a 2sec cast, the elemental never does anything 0.5sec after the cast is over..... so I have chosen to effectively model it as a 2.5sec cast
(3) Fire Blast also seems to have a 5sec CD
(4) There is no clear priority list. If both FB and FN are off cooldown, it seems random which one is picked
(5) Swing time is 2sec and un-modified by any haste effects
(6) If FB or FN will come off CD in less than 2sec, the Elemental will often just wait instead of trying to get in another melee swing. I haven't yet determined the crossover point yet.
(7) FN costs 207 mana and FB costs 276 mana
(8) The elemental has 4k base mana and inherits 30% of master's intellect, gaining 15 mana per intellect
(9) When out of the five-second-rule it gains 10.66 mana per second

Hope that helps. I'll have a version of SimC out (possibly tonight) with my implementation for comparison purposes if necessary.

Levva · 11/09/09, 11:48 AM

I was looking at the sim output trying to determine why Rawr.Enhance likes hit more than EnhSim likes hit. I noted the following phenomena. With both weapons having Berserking on them the MH Uptime was given as 42.18% and the OH as 37.23% however I was at a loss to think of a reason that the OH proc rate would be so much lower than the MH proc rate. Any thoughts/explanations?

Rouncer · 11/09/09, 12:12 PM

Originally Posted by Levva

I was looking at the sim output trying to determine why Rawr.Enhance likes hit more than EnhSim likes hit. I noted the following phenomena. With both weapons having Berserking on them the MH Uptime was given as 42.18% and the OH as 37.23% however I was at a loss to think of a reason that the OH proc rate would be so much lower than the MH proc rate. Any thoughts/explanations?

Windfury hits can proc Berzerking

Ryethe · 11/09/09, 12:12 PM

Originally Posted by Levva

I was looking at the sim output trying to determine why Rawr.Enhance likes hit more than EnhSim likes hit. I noted the following phenomena. With both weapons having Berserking on them the MH Uptime was given as 42.18% and the OH as 37.23% however I was at a loss to think of a reason that the OH proc rate would be so much lower than the MH proc rate. Any thoughts/explanations?

Were they both 2.6 speed?

EDIT: I did some testing and the uptime persisted regardless of weapon speed.

EDIT2:
I ran a pure white hit test and am really confused by the results:

EnhSim version 1.8.5.2

...............................................................................


MH:
miss                2427018        9.51%
dodge               2600           0.01%
glancing            6121848        24.00%
hit                 4683856        18.36%
crit                12275527       48.12%
average ap          8157.07
clip                0              0.00%

MH Windfury:
procs/hits          16.42%
procs/swings        14.85%
miss                0              0.00%
dodge               780            0.01%
hit                 3705013        48.89%
crit                3872245        51.10%
average ap          8160.90

OH:
miss                2336396        9.52%
dodge               2406           0.01%
glancing            5894269        24.01%
hit                 4507579        18.36%
crit                11808753       48.10%
average ap          8159.76
clip                0              0.00%

OH Flametongue:
miss                0              0.00%
hit                 12757259       57.44%
crit                9453342        42.56%
average sp          3047.38

Maelstrom Weapon:
PPM                 38.80

flurry uptime                  85.65%
unleashed rage uptime          99.97%
elemental devastation uptime   0.00%
elemental oath uptime          0.00%

mh enchant uptime              35.49%
oh enchant uptime              23.99%
trinket1 uptime                20.18%
trinket2 uptime                32.06%
racial uptime                  0.00%
totem uptime                   0.00%
set bonus uptime               0.00%
metagem uptime                 0.00%
gloves enchant uptime          0.00%
cloak enchant uptime           0.00%

                    MRPS without overregen
mana regen          0.00      22.66%
mp5                 0.00      165.93%
replenishment       -0.00     -50.18%
unrelenting storm   0.00      0.00%
judgement of wisdom -0.00     -38.41%
shamanistic rage    0.00      0.00%
mana potion         0.00      0.00%
misc                0.00      0.00%
stormstrike         0.00      0.00%

                    DPS                 PPM       MPS
white               2288.57   64.72%    83.43
windfury            810.17    22.91%    6.32
flametongue         437.59    12.37%    37.02

DPS                 3536.32
MPS                 0.00
MP2min              0.00
MRPS                93.52
Out of mana time    0.00%

elapsed simulation time: 10000.00h
elapsed real time: 18.69s
simulation speed: 1926472x

I remember reading about having 50% of the proc rate on off-hand weapons but I thought that was no longer the case? Also the OH number isn't nearly 50% of the MH number in this case so that can't be it. Correct me if I'm wrong but MH and OH should have the same hit%, right?

Stopokingme · 11/10/09, 5:34 AM

Afaik the 50% offhand proc rate is a relic from 1.0, from rogues testing, not taking into account the extra main hand strikes triggering extra procs. You can safely disregard that.

• Jessamy · 11/10/09, 9:50 AM

You did not do a pure white hit test. You still have windfury on your main hand and flametongue on your off hand. Since your test includes windfury attacks, the enchant uptimes will be skewed in your results. In-game (or in a more standard simulation run), stormstrike and lava lash will balance this some, since they add more off hand swings.

Here is a test run of the BiS thread config with no weapon imbues, and all attacks turned off. Enchant uptimes are very similar now.

EnhSim version 1.8.5.0

................................................................................

MH:
miss                1190567        8.73%
dodge               45892          0.34%
glancing            3274297        24.01%
hit                 2062187        15.12%
crit                7064564        51.80%
average ap          8375.71        
clip                0              0.00%

OH:
miss                1191761        8.74%
dodge               46366          0.34%
glancing            3273344        24.00%
hit                 2073985        15.21%
crit                7052053        51.71%
average ap          8361.05        
clip                0              0.00%

Maelstrom Weapon:
PPM                 35.80

flurry uptime                  86.61%
unleashed rage uptime          0.00%
elemental devastation uptime   0.00%
elemental oath uptime          0.00%

mh enchant uptime              24.09%
oh enchant uptime              24.19%
trinket1 uptime                20.07%
trinket2 uptime                32.02%
racial uptime                  0.00%
totem uptime                   0.00%
set bonus uptime               0.00%
metagem uptime                 0.00%
gloves enchant uptime          0.00%
cloak enchant uptime           0.00%

                    MRPS without overregen
mana regen          0.00      54.76%
mp5                 0.00      99.34%
replenishment       -0.00     -30.05%
unrelenting storm   0.00      0.00%
judgement of wisdom -0.00     -24.06%
shamanistic rage    0.00      0.00%
mana potion         0.00      0.00%
misc                0.00      0.00%
stormstrike         0.00      0.00%

                    DPS                 PPM       MPS
white               2780.68   100%      90.92

DPS                 2780.68
MPS                 0.00
MP2min              0.00
MRPS                97.74
Out of mana time    0.00%

elapsed simulation time: 5000.00h
elapsed real time: 13.97s
simulation speed: 1288659x

Raut · 11/11/09, 7:49 AM

Levva, any chance you could update the default ep_* values in the default config file to reflect actual item budget? It doesn't matter per se as it's just used to calculate EP but with values like 40 AP, 20 haste, etc you'll have something closer to actual gear changes. Given an empty socket it's even a perfect simulation of how to gem an item. Default AP and SP EP values are just so much higher than everything else and I don't see why.

Murmor · 11/11/09, 8:17 AM

I've spent some time trying to get up to speed with Enhsim and Rawr, to try and help figure out where they disagree.

One difference is that Rawr uses the same average AP for all abilities, for Enhsim they come out different. ToQE accounts for a lot of that, as the lava lash that procs it doesn't get the effect. I can't prove that at the moment, though.

Also, Enhsim can proc something off a main hand Stormstrike and have that effect apply to the offhand. Offhand SS is showing more average AP then mainhand SS. Does this happen in the game too?

Levva · 11/11/09, 10:37 AM

Originally Posted by Murmor

I've spent some time trying to get up to speed with Enhsim and Rawr, to try and help figure out where they disagree.

One difference is that Rawr uses the same average AP for all abilities, for Enhsim they come out different. ToQE accounts for a lot of that, as the lava lash that procs it doesn't get the effect. I can't prove that at the moment, though.

Also, Enhsim can proc something off a main hand Stormstrike and have that effect apply to the offhand. Offhand SS is showing more average AP then mainhand SS. Does this happen in the game too?

I'd be interested to try an EnhSim with say 100,000 hours or more time to see if the average AP & SP settled down on each ability. I wouldn't believe the current variations at present as the sim length is too short to see them settle down.

This should at least smooth out the effects of ToQE you describe as in theory as the combat sim length tends to infinity so the variations should tend to zero and work out to be the average. This is of course unless there is a complex feedback mechanism such as with WF, MW5, ED & haste, where different rules apply and even Rawr needs an iterative approach.

On default config.txt - yup sure I can change them - Do we reckon that it should be just the value of 1 epic gem ie: 40 AP, 23 SP and 20 everything else?

Raut · 11/11/09, 10:51 AM

Originally Posted by Levva

On default config.txt - yup sure I can change them - Do we reckon that it should be just the value of 1 epic gem ie: 40 AP, 23 SP and 20 everything else?

Not sure. 40/20/23 are good because they are one gem but slight deviation because of RNG are more significant. Perhaps 80/40/46?

Murmor · 11/11/09, 11:15 AM

Originally Posted by Levva

I'd be interested to try an EnhSim with say 100,000 hours or more time to see if the average AP & SP settled down on each ability. I wouldn't believe the current variations at present as the sim length is too short to see them settle down.

This should at least smooth out the effects of ToQE you describe as in theory as the combat sim length tends to infinity so the variations should tend to zero and work out to be the average. This is of course unless there is a complex feedback mechanism such as with WF, MW5, ED & haste, where different rules apply and even Rawr needs an iterative approach.

I think LL will always have a lower average AP than the other abilities, because of it's interaction with ToQE. During a fight you have time when Volcanic Fury is active and times when it isn't. You always know the last ability before it becomes active is LL. I'll run a 100,000 hour sim to prove it, though.

Rouncer · 11/11/09, 12:30 PM

How sure are you that the Lava Lash that procs Volcanic Fury isn't affected by it? I would work on testing that conclusion in game before worrying about how that is modeled in the sim.

Shouldn't be hard to test. Stand sideways to a level 60 dummy naked with a white/grey slow offhand (to minimize the damage range and maximize the effect of the proc) and start Lava Lashing. Once you have a decent control sample, equip the totem and do a bunch tracking only the ones that proc and then clicking it off. If there is no difference in the average value of the ones that proc versus the control group then you have proved your theory.

Murmor · 11/11/09, 1:14 PM

Originally Posted by Rouncer

How sure are you that the Lava Lash that procs Volcanic Fury isn't affected by it? I would work on testing that conclusion in game before worrying about how that is modeled in the sim.

Shouldn't be hard to test. Stand sideways to a level 60 dummy naked with a white/grey slow offhand (to minimize the damage range and maximize the effect of the proc) and start Lava Lashing. Once you have a decent control sample, equip the totem and do a bunch tracking only the ones that proc and then clicking it off. If there is no difference in the average value of the ones that proc versus the control group then you have proved your theory.

Fair point. I did this test and it didn't take long to conclude that LL works the way I thought. For the record with only [Hammer] in the offhand LL hits for 70-77 regardless of whether ToQE procs or not. Once the proc is active it will hit for about 110-120. It was pretty obvious that nothing more detailed was required.

With that in mind - the numbers for the 100,000 hour sim are:

Average AP for ability:

MH	7904.86
MH WF	7908.41
MS SS	7893.52
OH	7908.22
OS SS	7936.85
LL	7880.01

And the full output was

C:\Users\loz\Documents\src\EnhSim-30240\Release>EnhSim.exe murmor_config.txt
EnhSim version 1.8.5.2

................................................................................


MH:
miss                9569330        4.07%
dodge               0              0.00%
glancing            56452751       24.00%
hit                 53670936       22.82%
crit                115480121      49.10%
average ap          7904.86
clip                0              0.00%

MH Windfury:
procs/hits          16.37%
procs/swings        15.80%
miss                0              0.00%
dodge               0              0.00%
hit                 41441974       47.91%
crit                45056958       52.09%
average ap          7908.41

MH Storsmtrike:
miss                0              0.00%
dodge               0              0.00%
hit                 18502635       47.99%
crit                20053607       52.01%
average ap          7893.52

OH:
miss                9955551        4.07%
dodge               0              0.00%
glancing            58680297       24.00%
hit                 55742432       22.80%
crit                120141566      49.13%
average ap          7908.22
clip                0              0.00%

OH Flametongue:
miss                0              0.00%
hit                 191516379      60.53%
crit                124887936      39.47%
average sp          2973.16

OH Stormstrike:
miss                0              0.00%
dodge               0              0.00%
hit                 18467924       47.90%
crit                20088318       52.10%
average ap          7936.85

Earth Shock:
miss                0              0.00%
hit                 29061539       60.53%
crit                18949065       39.47%
average sp          2973.66

Lava Lash:
miss                0              0.00%
dodge               0              0.00%
hit                 20630958       47.66%
crit                22652820       52.34%
average ap          7880.01

Lightning Bolt:
miss                0              0.00%
hit                 31854266       60.52%
crit                20779391       39.48%
average sp          2978.96

Lightning Shield:
miss                0              0.00%
hit                 61874962       100.00%
crit                0              0.00%
average sp          2982.48

Spirit Wolves:
miss                0              0.00%
dodge               0              0.00%
glancing            40508300       24.00%
hit                 108881100      64.51%
crit                19392856       11.49%
average ap          7917.94

Magma Totem:
miss                0              0.00%
hit                 88767677       60.53%
crit                57873248       39.47%
average sp          2975.64

Maelstrom Weapon:
PPM                 49.15

flurry uptime                  90.68%
unleashed rage uptime          100.00%
elemental devastation uptime   74.05%
elemental oath uptime          0.00%

mh enchant uptime              43.69%
oh enchant uptime              37.68%
trinket1 uptime                32.46%
trinket2 uptime                18.07%
racial uptime                  0.00%
totem uptime                   91.40%
set bonus uptime               0.00%
metagem uptime                 0.00%
gloves enchant uptime          0.00%
cloak enchant uptime           0.00%

                    MRPS without overregen
mana regen          0.00      0.00%
mp5                 18.03     8.71%
replenishment       38.47     18.58%
unrelenting storm   0.00      0.00%
judgement of wisdom 8.96      4.33%
shamanistic rage    47.81     23.10%
mana potion         0.00      0.00%
misc                0.00      0.00%
stormstrike         93.74     45.28%

                    DPS                 PPM       MPS
white               2240.81   31.31%    79.95
windfury            947.24    13.23%    7.21
flametongue         577.49    8.07%     52.73
stormstrike         453.65    6.34%     6.43      35.41     16.03%
lava lash           369.12    5.16%     7.21      19.87     9.00%
magma totem         506.65    7.08%     2.49      46.29     20.96%
earth shock         585.70    8.18%     8.00      51.71     23.42%
lightning bolt      919.67    12.85%    8.77      64.27     29.10%
lightning shield    336.23    4.7%      1.15      0.00      0.00%
spirit wolves       220.80    3.08%     0.40      3.28      1.49%

DPS                 7157.36
MPS                 220.83
MP2min              26499.57
MRPS                214.56
Out of mana time    0.00%

elapsed simulation time: 100000.00h
elapsed real time: 560.74s
simulation speed: 642011x

Config File:

#################	
##ENHSIM CONFIG##	//comment marks are # and //
##  v1.8.5.2   ##   //if you write comments on the same line with config line
#################	//leave at least one space between the comment and value


//IMPORTANT: Be careful when inputting your stats. Read is the value rating or percent etc. Read the comments.
//You always input what you see on your character sheet/paper doll.

simulation_time			100000 	//in hours, you can also input decimals ex. 0.05
simulation_time_combatlog	300	//in minutes
combat_length			10	//in minutes
combat_length_rng_factor        0.1	//Maximum random adjustment to add to each combat run, specified as a decimal % (0.1 = 10%)
report_count			80 	//how many times the progress is reported during the simulation
threads				2	//number of threads to use, combatlogging always uses 1 thread
min_lag				250	//in milliseconds, lag means ping + player reaction time
max_lag				250	//1000 ms = 1s
simulate_mana			1	//{ 0 | 1 } if disabled, increases speed, but doesn't take mana limitations into account



########### //To understand the EP ranges you have to understand how EP values are calculated. 
#EP RANGES# //For example, to calculate EP value of crit rating, first I simulate the normal "baseline" dps.
########### //Then I simulate dps with increased crit rating. I take the dps difference and divide it by the
//increased crit rating to get dps/crit rating value. EP range means how much crit rating did I increase.

//Watch the expertise and hit caps. For example if expertise goes over the cap resulting from too high EP
//range, the EP value will go lower than it should be. On the other hand, if you give too low EP ranges,
//you won't get accurate results. Best scenario is low EP ranges and high simulation time.

//If hit/spellhit or expertise is already over the cap, test will be skipped. Although you can
//give negative ep range to see how much EP is hit/expertise worth to have. To skip a test, use 0 as range.

//Which stat is used as baseline stat, meaning it gets EP value of 1. This is important when calculating EP values.
ep_precision			2 //how many digits there are in ep value?
ep_base_stat			ap //{ ap | sp }
ep_ap 				200
ep_crit_rating 			30
ep_hit_rating 			30
ep_expertise 			4 //not rating
ep_haste_rating 		30
ep_armor_penetration_rating 	30
ep_spellpower 			150
ep_dps				7.5
ep_mana				250
ep_spirit			150
ep_mp5				20



###################
#ROTATION AND MISC#
###################
mh_auto_attack			1	//{ 0 | 1 }
oh_auto_attack			1	//{ 0 | 1 }
wait_ss_with_wf_cd		0.00	//max seconds to wait to SS, if wf cd is still up
cast_ll_only_if_wf_on_cd	0	//{ 0 | 1 } this is noted only if you have dual wf
bloodlust_casters		1	//{ 0 | 1 }
sync_bloodlust_with_trinkets	1	//{ 0 | 1 }
cast_lvb_only_if_ed_left	15.0	//cast LvB only if ED duration x or less seconds left
cast_lvb_only_if_fsdots_left	4	//maximum Flame Shock dot ticks left to cast LvB
cast_lvb_only_if_fs_active	1	//{ 0 | 1 }
cast_fs_only_if_dots_left	0	//maximum Flame Shock dot ticks left to cast Flame Shock
cast_ls_only_if_charges_left	0	//maximum Lightning Shield charges left to cast LS
cast_sr_only_if_mana_left	1250	//maximum mana left to cast SR
use_mana_potion_if_mana_left	3000	//maximum mana left to use mana potion


//set the rotation priority
//SW = spirit wolves
//SS = stormstrike
//FS = flame shock
//ES = earth shock
//ES_SS = earth shock only if there are stormstrike charges
//Frost = frost shock
//MW5_LB = maelstrom weapon cast lightning bolt, min 5 charges
//MW3_CL = maelstrom weapon cast chain lightning, min 3 charges
//generic form: MWx_YY, x is 1-5 and YY is either LB or CL
//LS = lightning shield
//LL = lava lash
//LB = lightning bolt
//CL = chain lighting
//LVB = lava burst
//ST = searing totem
//MT = magma totem
//SR = shamanistic rage
rotation_priority_count		8
rotation_priority1              SW
rotation_priority2		SR
rotation_priority3		MW5_LB
rotation_priority4		SS
rotation_priority5		ES
rotation_priority6		LL
rotation_priority7		MT
rotation_priority8		LS



##########
#BOSS#####
##########
miss				8.00	//%
dodge				6.50	//%
glancing			24.00	//%
armor				10643	
spell_miss			17.00	//%

nature_resistance		0	//effective resistances
fire_resistance			0
frost_resistance		0
arcane_resistance		0
shadow_resistance		0



###############################################################################
### Everything in the section below can be replaced by information obtained ###
### from your paper doll stats or exported by the ShockAndAwe addon         ###
###############################################################################


##########

##########################################
### Rawr.Enhance Data Export to EnhSim ###
##########################################

config_source rawr

race                            tauren
mh_speed                        2.7
oh_speed                        2.6
mh_dps                          196.5
oh_dps                          196.5
mh_crit                         33.08
oh_crit                         33.08
mh_hit                          16.93
oh_hit                          16.93
mh_expertise_rating             144
oh_expertise_rating             144
ap                              4373
melee_haste                     12.37
armor_penetration               7.57
str                             145
agi                             1038
int                             578
spi                             165
spellpower                      1312
spell_crit                      23.37
spell_hit                       21.16
spell_haste                     9.52
max_mana                        12786
mp5                             0

mh_imbue                        windfury
oh_imbue                        flametongue

mh_enchant                      3789
mh_weapon                       axe
oh_enchant                      3789
oh_weapon                       axe

trinket1                        47303
trinket2                        45286

totem                           47667
set_bonus1                      worldbreaker_battlegear_2
set_bonus2                      t9_battlegear_2
set_bonus3                      -

metagem                         41285

gloves_enchant                  -
cloak_enchant                   -
glyph_major1                    flametongue_weapon
glyph_major2                    stormstrike
glyph_major3                    windfury_weapon

glyph_minor1                    -
glyph_minor2                    -
glyph_minor3                    -


cast_sr_only_if_mana_left 1250
replenishment             1
water_shield              0
mana_spring_totem         0
blessing_of_wisdom        1
judgement_of_wisdom       1


#############
## Talents ##
#############

ancestral_knowledge             2/5
improved_shields                3/3
mental_dexterity                3/3
shamanistic_focus               1/1
flurry                          5/5
elemental_weapons               3/3
unleashed_rage                  3/3
weapon_mastery                  3/3
dual_wield_specialization       3/3
mental_quickness                3/3
improved_stormstrike            2/2
static_shock                    3/3
maelstrom_weapon                5/5
convection                      0/5
concussion                      5/5
call_of_flame                   3/3
elemental_devastation           3/3
reverberation                   0/5
elemental_focus                 0/1
elemental_fury                  5/5
call_of_thunder                 0/1
unrelenting_storm               0/3
elemental_precision             0/3
lightning_mastery               0/5
elemental_oath                  0/2
lightning_overload              0/5
lava_flows                      0/3
storm_earth_and_fire            0/3
shamanism                       0/5

#########
# Buffs #
#########

armor_debuff_major              20.0/20.0
armor_debuff_minor              5.0/5.0
physical_vulnerability_debuff   4.0/4.0
melee_haste_buff                20.0/20.0
melee_crit_chance_buff          5.0/5.0
attack_power_buff_flat          687/688
attack_power_buff_multiplier    0.0/99.7
spell_haste_buff                5.0/5.0
spell_crit_chance_buff          5.0/5.0
spell_crit_chance_debuff        5.0/5.0
spell_damage_debuff             13.0/13.0
spellpower_buff                 280/280
spell_hit_chance_debuff         3.0/3.0
haste_buff                      3.0/3.0
percentage_damage_increase      3.0/3.0
crit_chance_debuff              3.0/3.0
stat_multiplier                 10.0/10.0
stat_add_buff                   51/52
agi_and_strength_buff           178/178
intellect_buff                  60/60
bloodlust_casters               1
flask_elixir                    flask_of_endless_rage
guardian_elixir                 -
potion                          potion_of_speed
food                            fish_feast

Levva · 11/12/09, 8:31 AM

Originally Posted by Murmor

Fair point. I did this test and it didn't take long to conclude that LL works the way I thought. For the record with only [Hammer] in the offhand LL hits for 70-77 regardless of whether ToQE procs or not. Once the proc is active it will hit for about 110-120. It was pretty obvious that nothing more detailed was required.

With that in mind - the numbers for the 100,000 hour sim are:

Average AP for ability:

MH	7904.86
MH WF	7908.41
MS SS	7893.52
OH	7908.22
OS SS	7936.85
LL	7880.01

And the full output was

Interesting. I can look at amending Rawr.Enhance to modify LL effects. The other thing of note is the variance between MH SS & OH SS as these two are the only ones significantly off the 7908 base. Interesting MH is lower than base OH is higher. Any theories as to why?

stassart · 11/12/09, 8:43 AM

Originally Posted by Levva

Interesting MH is lower than base OH is higher. Any theories as to why?

[Formula: Enchant Weapon - Berserking] would be my theory. I have not tested if it is the behavior in game, but I am guessing that in EnhSim if Berserking procs on the MH hit of Stormstrike then the OH hit gets the benefit. A trinket ([Blood of the Old God] for example) that increases attack power and procs off the main hand hit could also be a source.

In game on Stormstrike, the main hand does hit before the off-hand, that is not difficult to verify and has been tested. In the case of windfury proc at least the proc is applied before the off-hand swing (in previous testing if windfury is applied to both weapons the MH will have a higher proc rate than the OH). I do not know if that means that other types of procs are applied before the OH hit though.

Edit: If the OH SS AP was actually higher in game, wouldn't the OH SS damage before more than half the MH SS damage? I just spent some time looking over WoL parses and in every large combatlog the OH SS was 50% of MH SS +- 1%. That does not seem to match what I would expect if the EnhSim behavior was correct? This is far from conclusive as 20 or so hours of in game combat comes nowhere near to the 5,000-10,000 hours of EnhSim runs, but it does make me question.

Murmor · 11/12/09, 11:06 AM

I ran some simple tests in game, with a hammer in one hand and a weapon with berserking on in the other to see how Stormstrike works.

What I concluded is that:

1. Either both MH and OH hit (either can be a crit), or SS as a whole is parried/dodged/missed. To me, it looks like Enhsim can have OH miss/parry/dodge independently of MH hit.
2. Berserking proc on MH nearly always affects the OH. I saw one occasion when it didn't. As hard as I stared at it, the number didn't change.
3. UR procs off the MH don't affect the OH

What I couldn't answer without a much longer test, (or for the rich, putting berserking on a hammer). Is whether the ability that procs berserking gets the AP bonus or not. My numbers are suggestive that it doesn't, but not conclusive.

Levva · 11/12/09, 12:17 PM

Originally Posted by Murmor

I ran some simple tests in game, with a hammer in one hand and a weapon with berserking on in the other to see how Stormstrike works.

What I concluded is that:

1. Either both MH and OH hit (either can be a crit), or SS as a whole is parried/dodged/missed. To me, it looks like Enhsim can have OH miss/parry/dodge independently of MH hit.
2. Berserking proc on MH nearly always affects the OH. I saw one occasion when it didn't. As hard as I stared at it, the number didn't change.
3. UR procs off the MH don't affect the OH

What I couldn't answer without a much longer test, (or for the rich, putting berserking on a hammer). Is whether the ability that procs berserking gets the AP bonus or not. My numbers are suggestive that it doesn't, but not conclusive.

You are always rich on the PTR. It is effectively fake money on the PTR assuming you transfer a character with heaps of cash. You can blow it all on berserking on a hammer and it won't effect your live character. This is probably worth doing as if EnhSim is inaccurately modelling Stormstrike or berserking procs then that needs fixed.

Levva · 11/13/09, 5:08 AM

Just to let people know I've had two individuals put themselves forward to assist with adding the T10 stuff to EnhSim, Eralun and Arbreng. Eralun will help with general coding and features, Arbreng is keen to get the multi-threading working again to dramatically speed up the time it takes the sim to complete.

Torcha · 11/13/09, 1:53 PM

(deleted)

Levva · 11/16/09, 5:28 AM

Further update...

Ziff has also now joined the EnhSim team and will be checking in his T10 patch changes soon. So hopefully we can have v1.9.0 ready BEFORE patch 3.3 hits.

Funaria · 11/17/09, 1:40 AM

I do not have logs, and am not sure how to link talents from the PTR if it is possible. Sorry for that. However, I haven't seen a reply yet to Rouncer's question as to whether Call of Flame affects the new Fire Nova spell.

The tooltip on the talent on the PTR says: Increases the damage done by your fire totems and Fire Nova by 5/10/15% and the damage done by your Lava Burst spell by 2/4/6%.

To ensure that it was actually working as the tooltip made it appear, I did a statistically insignificant number of tests (because I knew I couldn't show logs and was only checking to make sure there was no bug). First I tested on a target dummy with 5 points spent in Convection and no other talent points spent. Normal hits were between 942 and 980, with a crit of 1444. Damage went up gradually with each point spent in Call of Flame, culminating in a non-crit average of 1126 and a crit of 1771.

I was using a Frost Resistance totem to avoid any effects from the totem itself, and wore the same gear for all testing. (I didn't have the bag space to go naked.)

Levva · 11/21/09, 11:21 AM

I've just added Elam author of various articles & guides on EnhSim to the EnhSim project as a documentation editor. He will take on the task of updating the wiki on the EnhSim website. He should be able to start soon on fleshing out the wiki with various articles and compiling some proper documentation, and how to guides.

Extenze · 11/22/09, 1:18 AM

I doubt this deserves it's own thread, and since it's dependent on EnhSim I'll just post it here. Included is the compiled EXE and the source AU3, it's an AutoIt Script file.

[e] Ha, I forgot the program's description. It's the program Levva was talking about that will run a bunch of simulations, change the RNG factor's in the config files, and save the results into a CSV file. This is also Windows ONLY.

http://dl3.dropbox.com/u/816613/rngOmatic.rar

It's operation's pretty straight forward. You type in the value which combat_length_rng_factor will incriment ex: 0.05,0.1,0.15... until it hits it's maximum, defaulted to 1. You'll want to hit "Create Configs" before you start it. It'll look at the original config in the 3rd input box, copy it and modify it and create as many as you want until it hits the maximum RNG factor. Just drop this in your enhsim directory and it's good to go.

Then hit Start and it'll run enhsim with the appropriate config file, and parse the outputs automatically and dump it to the rngOutput.txt everytime the RNG factor changes, or every 10 sims.

This script will output the results into a text file, but it's format should be compatible with any spreadsheet software. Just change the filetype to *.csv

Sample output:
rng,dps1,dps2,...,dps10
0.05,3065.08,3065.02,3065.73,3065.06,3065.35,3065.32,3064.95,3064.80,3065.32,3064.34
0.1,3064.57,3064.46,3065.00,3064.96,3065.44,3064.58,3064.63,3065.67,3064.42,3064.36

If you want to stop it while it's working, you'll have to right click the tray icon and exit that way.

There's no INI support, but it didn't seem like a big deal because there's only 3 inputs and it'll take quite a while for a full run.

You shouldn't have to worry about enhsim updates breaking this, as long as you don't change the names of the DPS result and combat_length_rng_factor.

**note: This script is using a quick and dirty way of monitoring CPU usage, which it uses as a notification that it's finished simming. It takes the cpu% off the status bar. You can minimize taskmgr.exe if you want to. Just don't close it. The default cpu% threshold is 50%

Let me know if any extra features are needed like being able to change the # of sims per RNG or something. And of course if you find any bugs.

[top]Change Log

Revision 8
*Added: EJ Table formatting, It's now completely copy-paste friendly. You will have to manually add STDEV values, however.
I did include that column in the header.
*Added: Hours to time estimation
*More tuning

Revision 7
*Added: Added a file open dialogue to find config.txt easier
*Fixed: fixed derivative configs not updating combat_length_rng_factor

Revision 6
*Quite a large speed increase and is way more reliable now (15-20mins off a standard run)

Revision 5
*Added: rngOutput.txt now has the incriment and maximum in it's name eg: rngOut_0.05_to_1.txt
*Added: sims per RNG, this lets you alter how many passes it makes per incriment, default 10
*Added: Add a starting point, so you can resume. Things happen

Revision 4
*Fixed the sim loop only working with 0.05
*It now starts with a config at 0 RNG factor

Revision 3
*Decided versions aren't my style
*Further corrections to estimated time

Version 1.0.1
*Added nopause too the end of execute string, nice time reduction
*Fixed time estimation
*No more includes, pulled everything into script. Reduced it's size by 2/3

Version 1
*Initial release