Glacial Spike - SimC vs Live

[Chuk]

Hello everyone!

I'm doing this thread because i'm not sure if SimC is misscalculating GS damage vs live or if there's a GS ingame bug.

I'm currently using this APL for testing ingame vs simc results.

Code: Select all

mage="Chukeh" level=110 race=undead region=us server=azralon role=attack professions=tailoring=708/enchanting=734 talents=2203122 spec=frost artifact=53:0:0:0:0:783:1:784:3:785:3:786:3:787:3:788:3:789:2:791:3:794:1:797:1:798:1:1296:1 head=,id=134425,bonus_id=3415/1522/3337 neck=,id=137311,enchant_id=5439,bonus_id=3414/1808/1507/1813,gem_id=130222 shoulder=,id=139191,bonus_id=1806/1517/3337 back=,id=142215,enchant_id=5433,bonus_id=3453/1482/3336 chest=,id=139193,bonus_id=1805/1487 wrist=,id=139130,bonus_id=3418/1542/3337 hands=,id=134261,bonus_id=3416/1522/1813 waist=,id=134215,bonus_id=3474/1808/1507/1674,gem_id=130219 legs=,id=134427,bonus_id=3417/1522/3336 feet=,id=139195,bonus_id=1805/1502/3337 finger1=,id=138220,enchant_id=5425,bonus_id=1805/1487 finger2=Lendario,stats=1850crit_150mastery_650haste trinket1=,id=134204,bonus_id=3473/605/1507/3336 trinket2=,id=141482,bonus_id=1487/3337 main_hand=,id=128862,bonus_id=731,relic_id=1806:1502/1807:1477:3336/1805:1487,gem_id=139250/139269/139259/0 # This default action priority list is automatically created based on your character. # It is a attempt to provide you with a action list that is both simple and practicable, # while resulting in a meaningful and good simulation. It may not result in the absolutely highest possible dps. # Feel free to edit, adapt and improve it to your own needs. # SimulationCraft is always looking for updates and improvements to the default action lists. # Executed before combat begins. Accepts non-harmful actions only. #actions.precombat=flask,type=flask_of_the_whispered_pact #actions.precombat+=/food,type=nightborne_delicacy_platter #actions.precombat+=/augmentation,type=defiled actions.precombat+=/water_elemental actions.precombat+=/snapshot_stats actions.precombat+=/mirror_image #actions.precombat+=/potion,name=potion_of_prolonged_power actions.precombat+=/frostbolt # Executed every time the actor is available. actions=counterspell,if=target.debuff.casting.react actions+=/time_warp,if=(time=0&buff.bloodlust.down)|(buff.bloodlust.down&equipped.132410) actions+=/call_action_list,name=cooldowns actions+=/flurry,if=buff.brain_freeze.react&prev_gcd.frostbolt actions+=/blizzard,if=buff.zannesu_journey.stack>4|buff.zannesu_journey.remains<cast_time+1&equipped.133970 actions+=/ice_lance,if=buff.fingers_of_frost.react>0&buff.chain_reaction.react>2&(equipped.138140|equipped.132411) actions+=/ice_lance,if=buff.fingers_of_frost.react>0&buff.chain_reaction.react>1&buff.rune_of_power.down actions+=/frozen_orb,if=buff.rune_of_power.down actions+=/glacial_spike #actions+=/frozen_orb actions+=/frostbolt actions.cooldowns=rune_of_power,if=cooldown.icy_veins.remains<cast_time|charges_fractional>1.7&cooldown.icy_veins.remains>38|buff.icy_veins.up|target.time_to_die.remains+5<charges_fractional*22 #actions.cooldowns+=/icy_veins,if=buff.icy_veins.down #actions.cooldowns+=/mirror_image actions.cooldowns+=/blood_fury actions.cooldowns+=/berserking actions.cooldowns+=/arcane_torrent #actions.cooldowns+=/potion,name=potion_of_prolonged_power,if=cooldown.icy_veins.remains<5

Things to consider in this APL

- 3rd tier is disabled. I wanna check GS damage only so i "muted" the tier so nothing can keep changing the damage of this skill.
- I created a similar item do the selphuz secret, just to get rid of the proc.
- FoF is never used.
- No consumables are used, Icy veins is not being used as well. Bloodlust is the only cooldown being used.
- Frostbolt damage and Flurry damage are absolutely on point.
- Rotation is FB X5 (or till 5 icicles) > GS

So, this is the results that a 5 min simulation.

Code: Select all

Simulation Iteration Average Direct Results Count Pct Mean Min Max Mean Min Max Actual Amount Total Amount Overkill % hit 20.83 73.86% 1128690.06 747162 2154384 1129532.96 978940 1339368 23515865 23515865 0.00 crit 7.37 26.14% 2257801.91 1494324 4308769 2260861.22 1757005 3558250 16650542 16650542 0.00

But in game, my results are usually like this.

Code: Select all

Hits - Min = ~800kk / Mean = ~1.1kk / Max = ~1.5kk Crits - Min = = ~1.5kk / Mean = ~1.85kk / Max = ~2.2kk

As you can see, although the mins are fine, the Max amounts are WAY off, on both hits and criticals. This usually means that on a 5 min fight, simcraft overvalues GS almost 10% more than live servers.

So, what i need to understand here is that if this is a live bug, a simcraft bug or maybe i'm missing something here? Because i honestly can't understand at all why this is happening.

All the tests are being done on the Raid training dummy on the Mage Class hall.

If you can, please help me test this as well.

[Chuk]

After quick discussion on MIRC, i found that Black Ice can be the reason why SimC is overvaluing GS.

Check the numbers below, removing black ice trait from my weapon:

Code: Select all

Simulation Iteration Average Direct Results Count Pct Mean Min Max Mean Min Max Actual Amount Total Amount Overkill % hit 20.86 73.96% 986699.74 747162 1216236 987256.16 870997 1103658 20587200 20587200 0.00 crit 7.34 26.04% 1972820.09 1494324 2432471 1973646.01 0 2432471 14490215 14490215 0.00

[hellfreeze]

Your max crits should be double your max hits, it means your sample was too small and you crit with smaller GS than you could have.

[Chuk]

It was pointed out to me that by not properly shattering FBs my results could be off.

So to make sure that shattering wasn't the issue, i removed flurry from the APL, and lust as well.

Here are the results:

Code: Select all

Type Executes Direct Results Ticks Tick Results Execute Time per Execution Tick Time per Tick Actual Amount Total Amount Overkill % Amount per Total Time Amount per Total Execute Time damage 31.58 31.55 0.00 0.00 2.3483 0.0000 40227686.97 40227686.97 0.00 542517.69 542517.69 Simulation Iteration Average Direct Results Count Pct Mean Min Max Mean Min Max Actual Amount Total Amount Overkill % hit 23.37 74.07% 1012783.81 747162 1684639 1013280.85 896564 1174108 23663781 23663781 0.00 crit 8.18 25.93% 2024913.12 1494324 3369278 2026297.27 1494324 2724279 16563906 16563906 0.00

I've also logged some of my tests. Here's the link:

https://www.warcraftlogs.com/reports/Aa ... ne&fight=2" target="_blank

As you can see, crits are still off by 10%.

Let me know if you need further testing.

[Chuk]

Another thing to mention:

By removing the flurry shatter from the APL (&prev_gcd.frostbolt) the DPS loss is about 1%. Which doesn't explain how the averages and criticals max are so far off.

[Norrinir]

Another thing to mention:

By removing the flurry shatter from the APL (&prev_gcd.frostbolt) the DPS loss is about 1%. Which doesn't explain how the averages and criticals max are so far off.

This doesn't remove Flurry shatter, it will still happen quite often. Try adding

Code: Select all

distance=1

to the profile.

[Rikx]

Since the whole point is finding out whether GS is being miscalculated, why not just remove flurry from the calculation?

[Norrinir]

Did some testing against a target dummy, apart from this bug, the damage is precisely where it should be.

30 hits, average damage 420463. SimC average 424233. Minimum and maximum won't match exactly, because the sample size is so small. To get the maximum, you need 5 Frostbolt crits and 5 Black Ice procs.

[Chuk]

Since the whole point is finding out whether GS is being miscalculated, why not just remove flurry from the calculation?

i did that, check my post above yours

[Mage]

I did two full nights of EN logs and results are on par with what simcraft shows, doesn't seem buggy.
Only the RNG of 1try per boss (on farm) is beyond frustrating.

[Rikx]

Since the whole point is finding out whether GS is being miscalculated, why not just remove flurry from the calculation?

i did that, check my post above yours

I meant not just removing the &prev_gcd.frostbolt section, but removing the line with Flurry in it in total.