Jump to content


Photo

Mists mechanics testing


  • Please log in to reply
74 replies to this topic

#21 Torima

Torima

    Von Kaiser

  • Members
  • 32 posts

Posted 06 August 2012 - 12:46 AM

From the looks of it poison's are applying at the amount on the tool tip with 50% chance per hit for assassination and 30% chance per hit for combat and at the same rate between weapon speeds and haste rating.

Assassination
1.8/1.8 speed 15.78%/2208 haste 50.4% proc/hit
World of Logs - Real Time Raid Analysis

2.6/1.8 speed 15.78%/2208 haste 52.898% proc/hit
World of Logs - Real Time Raid Analysis

Combat
1.8/1.8 speed 15.78%/2208 haste 31.20% proc/hit
World of Logs - Real Time Raid Analysis

2.6/1.8 speed 15.78%/2208 haste 30.19% proc/hit
World of Logs - Real Time Raid Analysis

1.8/1.8 speed 17.90%/3053 haste 31.15% proc/hit
World of Logs - Real Time Raid Analysis

#22 nextormento

nextormento

    Von Kaiser

  • Members
  • 93 posts

Posted 06 August 2012 - 01:32 AM

[Shadow Blades]


It'd be good to have some WoL or simply timestamped logs for this, yes:
Mainly to check if the procs occur exactly when a swing is expected, or if it has any kind of swing reset or server-client bugginess. Also, we've seen some missed procs being reported in combat logs as melee swing misses; perhaps that's worth investigating too.

#23 Isebel

Isebel

    Glass Joe

  • Members
  • 21 posts

Posted 06 August 2012 - 02:27 AM

40d

World of Logs - Real Time Raid Analysis

[19:03:00.569] Xarran Deadly Poison Raider's Training Dummy 5836
[19:03:01.031] Xarran Shadow Blade Raider's Training Dummy *33020*
[19:03:01.031] Xarran Shadow Blade Off-hand Raider's Training Dummy 11042
[19:03:01.031] Xarran Deadly Poison Raider's Training Dummy *22868*
[19:03:01.402] Xarran gains 15 energy from Xarran's Combat Potency
[19:03:01.408] Xarran Revealing Strike Raider's Training Dummy 18808

It appears that Shadow Blades can proc Combat Potency and Deadly Poison. I haven't tested for weapon enchant procs but seeing as Shadow Blades appears to work like a modified auto attack I wouldn't be surprised if that were the case as well.

Also 40a

World of Logs - Real Time Raid Analysis

Having attack speeds of 2.12 and 1.47 in MH and OH respectively, you can theoretically calculate you'd see 12/2.12 and 12/1.47 Shadow Blades procced. Seeing as most of the information I've seen so far it's to no surprised Shadow Blade doesn't really proc it's just an empowered auto attack as I had 6 Shadow Blades hit in each cast as well as 8 Shadow Blade Off-hands hit in each cast.

Now I'm not white hit cap, I am special hit cap though and I am yet to see a miss on a Shadow Blade in a heroic that I haven't logged nor any of those in the log above. I'd venture to say that Shadow Blade is an empowered auto attack and behaves the same as one BUT it is possible that it goes off of Special Hit cap. That could probably do for more iterations of tests but from what I've seen so far it appears they also go off Special Hit cap.

#24 Sakuratei

Sakuratei

    Piston Honda

  • Members
  • 188 posts

Posted 06 August 2012 - 02:10 PM

Wrapping up (15.), both the instant and DoT compontent of Deadly Poison scale with Bandit's GUile, as well as Wound Poison. The DoT component of Deadly Poison behaves the same way as Rupture and Crimson Tempest (the DOT doesn't update with Bandit's Guile until it is reapplied), although due to its rapid reapplication it will barely be noticed.

Edit: 40d. The picture linked here shows evidence of Main Gauche proccing from Shadow Blades.

#25 shadowboy813

shadowboy813

    Von Kaiser

  • Members
  • 50 posts

Posted 06 August 2012 - 03:33 PM

Regarding Elemental Force. It appears to proc as a spell using spell hit/crit mechancs as indicated in this log. That log was done at 7.63% hit,1.84% expertise (if I remember correctly), and 9.48% spell hit (all values taken from character sheet).

Another note relative to level 85 content. Throw and Deadly throw can both proc vial of shadows. Shurikan toss can as well, but it's not really relevant at level 85.

Dual wielding with Windsong: The other weapon basically just increases the uptime, offhand procs will refresh a mainhand proc (in other words, there was never more than 3 buffs up at once).

Dual wielding 1.8 daggers, autoattacks only, no poisons: log
Single wielding a 1.8 dagger, autoattacks only, no poisons: log

Both tests were run for the same total number of hits, noting the difference in uptime and similarity of number of procs.


Here are some questions not specifically asked that I investigated but may be important and/or interesting to know:

Combo point behavior: Premeditation does NOT add any anticipation charges. If used at 4 combo points, premed grants one combo point and the second does not overflow into an anticipation charge, it is merely wasted.

I also noted the timer on slice and dice to interpret how SnD behaves with respect to the "dot extend" mechanics of envenom and revealing strike. What I have found is that SnD follows that behavior with a 2sec tick timer. I used slice commander addon and noted buff timing. I found that if I refreshed with a 5pt SnD at 2.9 sec, it was extended to 36.9sec. If repeated at 1.9sec, it was extended to 37.9sec. This is not unreasonable given that SnD gives energy at 2sec intervals for subtlety (which was the spec I tested it in). It may be worth confirming if it behaves the same way for combat and assassination.

Edit number x: I tested the heal behavior of leeching poison and how it interacts with bandit's guile. The question I was wondering was if the leeching poison heal was derived from the base weapon swing before applying bandit's guile, or the final weapon strike damage after applying bandit's guile. The leeching poison heal is based on the weapon strike damage AFTER bandit's guile is applied (so basically you get 30% bigger leeching poison heals in deep insight than you do at no insight). Also wound poison is benefitting from bandit's guile. I have not tested deadly. There is also a slight delay from the weapon strike and the leeching poison heal it triggers.

I used an item level 4, 1.6 speed dagger with a damage range of 3-5. I unequipped all items that proc increased agility/ap.

Snipped log (for readability), with a single typical melee swing, leeching poison heal & wound poison proc per insight level:
Spoiler


Edit number x+1: Recuperate can proc windsong. I have not observed any procs from shiv, nonlethal poisons, or deadly poison dot ticks, still need more testing on poison proc behavior though.

#26 nextormento

nextormento

    Von Kaiser

  • Members
  • 93 posts

Posted 06 August 2012 - 07:52 PM

[Poisons and Bandit's Guile]


They seem to behave like expected. However, given that this is mechanics testing, I would like to know how you addressed this test. I'm not particularly concerned about snapshoting but, if someone can provide logs with highest-lowest DP-I procs under Bandit's Guile, that would be very appreciated.

#27 Torima

Torima

    Von Kaiser

  • Members
  • 32 posts

Posted 06 August 2012 - 11:38 PM

5.
Testing with H No'Kaled the damage proc is going off of the spell crit and appears to not use any hit table as it always hit no matter what.

Tested at level 85. 6.55% spell crit.
85 target dummy
World of Logs - Real Time Raid Analysis
120 procs 8 crits 0 misses 6.67% crit seen

93 target dummy
World of Logs - Real Time Raid Analysis
61 procs 0 crits 0 misses

Vial of Shadows which was physical going off of melee crit and always hit as well regardless of miss rate 16.41% crit.
85 target dummy
World of Logs - Real Time Raid Analysis
241 procs 34 crits 0 misses 14.1% crit seen

#28 Sakuratei

Sakuratei

    Piston Honda

  • Members
  • 188 posts

Posted 07 August 2012 - 05:27 AM

They seem to behave like expected. However, given that this is mechanics testing, I would like to know how you addressed this test. I'm not particularly concerned about snapshoting but, if someone can provide logs with highest-lowest DP-I procs under Bandit's Guile, that would be very appreciated.


No gear, weapon only, as with the other Bandit's Guile tests. DoT behaviour checked by observing tick damage in combat log when DI falls off.

http://img20.imagesh...340/test1po.jpg

#29 Enzo90910

Enzo90910

    Von Kaiser

  • • Guide Author
  • 89 posts

Posted 08 August 2012 - 10:54 AM

I also noted the timer on slice and dice to interpret how SnD behaves with respect to the "dot extend" mechanics of envenom and revealing strike. What I have found is that SnD follows that behavior with a 2sec tick timer. I used slice commander addon and noted buff timing. I found that if I refreshed with a 5pt SnD at 2.9 sec, it was extended to 36.9sec. If repeated at 1.9sec, it was extended to 37.9sec. This is not unreasonable given that SnD gives energy at 2sec intervals for subtlety (which was the spec I tested it in). It may be worth confirming if it behaves the same way for combat and assassination.


Just tested it for Assassination and it behaves the same, whether refreshed by applying SnD or refreshed with an Envenom. Got to aim for that 2s window to avoid losing energy on SnD (obviously it won't happen that often for Assasination rogues, who refresh SnD for free all the time).

#30 shadowboy813

shadowboy813

    Von Kaiser

  • Members
  • 50 posts

Posted 10 August 2012 - 02:09 PM

Question 13:

Revealing strike is NOT normalized
Main Gauche is still normalized.

I attacked a level 60 dummy using white gear (no stats), and an item level 4, 1.9 speed mace with a tooltip listed damage range of 3-6. The mace was 2.48 dps.

I ran several autoattacks on a level 60 target dummy to compare the damage to the character sheet (AP = 921) damage range of 128-131 to determine armor mitigation. The result was 119-121 damage, 120 average, for an armor factor of about 0.92.

The average damage of Main Gauche was 180.
Expected damage if normalized is: 1.2*0.92*(2.48*1.9 + 2.4*921/14) = ~179.5, very close to the average of 180.
If not normalized, expected damage is 143.

After repeating by casting revealing strike 4 times, waiting for shallow insight to fall off, and repeating, I got an average revealing strike damage of 180 (damage range of 179 to 182, only 13 casts, but the damage range is narrow enough for that to be conclusive).
Expected damage if normalized: 1.5*0.92*(2.48*1.9 + 2.4*921/14) = ~224,3. Average damage was 180, which is well outside that range.
Expected damage if not normalized: 1.5*0.92*(2.48*1.9 + 1.9*921/14) = 179, very close to the 180 observed.

Another noted observation. The revealing strike (and presumably sinister strike as well) that triggered advancement of bandit's guile did not benefit from the new bandit's guile state.

#31 Aldriana

Aldriana

    Mike Tyson

  • Moderators
  • 13,510 posts

Posted 22 August 2012 - 12:53 AM

I'd like to add one question to the list, namely: we were told here that poisons would use melee hit mechanics. Did this actually happen, and if so, what does it actually mean? Can poison procs now be dodged independently from the attack that triggered them? Can they glance? Do they use 1H hit rates or DW hit rates?

#32 Torima

Torima

    Von Kaiser

  • Members
  • 32 posts

Posted 22 August 2012 - 03:35 AM

As far as I have seen they are using the special hit cap and do not glance and can't be dodged if the original attack hit. Doing dps testing on the new dummies with only 7.7% hit and 2.2% exp with out a single miss and have yet to encounter a dodge World of Logs - Real Time Raid Analysis.

This is a log with both hit and expertise above the 7.5% cap World of Logs - Real Time Raid Analysis

This log is from when the dummies could parry and block from behind with 7.7% hit and 2.5%exp. World of Logs - Real Time Raid Analysis

I will need to make a log for being under the 7.5% hit cap to confirm if poisons will miss then but there is a post here from Madsushi at 85 hitting the 93 raiders dummy http://elitistjerks...._all_specs/p30/ and only getting misses.

#33 Isebel

Isebel

    Glass Joe

  • Members
  • 21 posts

Posted 22 August 2012 - 04:13 AM

I really wanted to see if Blizzard held their end on that statement in particular Aldriana, I have a whole lot of logs done as every spec with close to expertise and hit cap here World of Logs - Real Time Raid Analysis

That would take some time to look through but it's better than clogging up the post with 10+ linked logs. I had not seen any dodges misses or glances from poisons as any of the specs (no parries but I was actually trying to test damage so I was behind the boss). I was mostly looking at general breakdowns and average hits between Beta builds and not super in depth analysis so I could have missed something possibly so I'll look through them again myself.

EDIT: Not fully Exp and Hit cap, forgot I had swapped a few things couple days ago and had never re-reforged the first round of logs. Day 2 and Day 3 should be fully Hit and Exp capped.

I also noticed a few new things that I could just knock out for you guys through those logs in 43 and 44

To answer 43 Killing spree Proccing MG it appears so World of Logs - Real Time Raid Analysis

[22:07:36.601] Xarran Killing Spree Internal Test Boss *40467*
[22:07:36.601] Xarran Killing Spree Off-Hand Internal Test Boss 11339
[22:07:36.710] Xarran Main Gauche Internal Test Boss 20531

and for 44 World of Logs - Real Time Raid Analysis I tested Ambush with just 2 1.8 pvp daggers, just opening with an Ambush out of stealth and Shadow Dancing right after so the Ambushes have MoS with them, that shouldn't be too deceiving in this log though considering the difference is much more than a 10% difference. It appears that the original Ambush is not getting the the Find Weakness buff.

[00:32:06.587] Xarran Ambush Internal Test Boss 22789
[00:32:07.490] Internal Test Boss afflicted by Find Weakness from Xarran
[00:32:08.735] Xarran Ambush Internal Test Boss 29445
[00:32:09.519] Internal Test Boss's Find Weakness is refreshed by Xarran
[00:32:10.190] Xarran Ambush Internal Test Boss 29466
[00:32:10.999] Internal Test Boss's Find Weakness is refreshed by Xarran
[00:32:11.397] Xarran Ambush Internal Test Boss 31682
[00:32:12.150] Internal Test Boss's Find Weakness is refreshed by Xarran

#34 nextormento

nextormento

    Von Kaiser

  • Members
  • 93 posts

Posted 24 August 2012 - 01:18 AM

You may have noticed that neither wowhead or wowdb have the tooltips for MoP trinkets matching the ingame tooltips. Most of them are tagged as scaling spells in the game files, and there's a good chance that this is related to how will they scale down in challenge modes. Now that those are being tested would be a good opportunity to check how the tooltips change and see if we can figure how to extract the correct values. I don't know what's the availability of raid trinkets, but pvp ones (or the relic of xuen from the dmf) can be useful for this test too.

#35 Pathal

Pathal

    Don Flamenco

  • • Guide Author
  • 335 posts

Posted 25 August 2012 - 03:31 AM

Beta Class Balance Analysis Pt. II - Forums - World of Warcraft

Ask, and ye shall receive:
River's Song: 2PPM on melee attacks that land, or are dodged, or parried.
Dancing Steel: 1PPM on melee attacks that land.
Colossus: 3PPM on melee attacks that land, or are dodged, or parried, with a 3-second cooldown.
Elemental Force: 5PPM on melee damage, or non-periodic spell damage/healing, with a 0.1-second cooldown.
Jade Spirit: 10% chance on spell damage or healing, 50-second cooldown.
Windsong: 1PPM on melee damage, or non-periodic spell damage/healing, with a 1-second cooldown.



#36 shadowboy813

shadowboy813

    Von Kaiser

  • Members
  • 50 posts

Posted 28 August 2012 - 08:06 PM

I've been investigating Crimson Tempest on beta and I have found something that went counter to my expectation.

The DoT is apparently ticking every 2 sec, and It does significantly more than 30% of the initial damage now. I was seeing the total dot damage summing up to 240% of the original damage, meaning 6 ticks, each tick at 40% of the original damage.

In just a quick 1-CT with no proc trinkets test that I just did, the original 5-point CT hit for 5185. After that was a 12sec DoT that ticked 6 times for 2074 damage per tick (exactly 40% of the original damage per tick, so the bleed totaled 240%). In testing the overwrite mechanics, hitting CT at 3.6sec left on the DoT extended it to 13.6sec, meaning it now acts on a 2 second interval. This is with the mists beta build 16016.

Posted Image

#37 Athariel

Athariel

    Glass Joe

  • Members
  • 12 posts

Posted 28 August 2012 - 09:29 PM

More on Crimson Tempest;

It appears to double-dip from both Mastery and Sanguinary Vein for Subtlety.

Posted Image

Crimson Tempest did 6227 damage. If we follow the assumption the poster above said, of 40% of the initial attack per tick, we would have 6227*0.4 = 2490.8.

However, we had 4503 instead, which is 4503/6227 = 0,723141159.

Lets add in Mastery and Sanguinary Vein to the 40% value (with 50.66% (1593 rating) mastery at 85 on the beta):
0,4*1.5066*1.2 = 0.723168
or
2490.8*1.5066*1.2 = 4503.167136

Which I would say, considering the rounding of the mastery, is rather spot on.

#38 shadowboy813

shadowboy813

    Von Kaiser

  • Members
  • 50 posts

Posted 28 August 2012 - 10:15 PM

I decided to look further into dot refresh mechanics--in particular rupture.

As we all know, if you apply rupture with <2 sec left, it adds the remainder to a new 24sec DoT. What I wanted to know is how the 12th tick of the old rupture's damage is calculated. The last tick of the old rupture is applied with the damage of the new rupture, which means if you clip 1 tick of rupture, you get 11 ticks at the old value and 13 ticks of the new value. This happens whether or not the new one is stronger or weaker than the old one.

I tested as combat in statless gear, first using a no-insight, non-revealing strike rupture, and clipping the last tick with a revealing strike deep insight rupture. I then repeated but in the opposite order. In both cases there were 11 ticks of the first rupture and 13 ticks of the new rupture.

Low to High:
Posted Image

High to Low:
Posted Image

Implication? For a combat rogue to clip a deep insight rupture would be a damage loss. Conversely it's ALWAYS worthwhile to clip the last tick a lower damage rupture with a higher damage one.

It may be worth testing if CT and maybe garrote have the same behavior.

#39 Fae

Fae

    Von Kaiser

  • Members
  • 54 posts

Posted 29 August 2012 - 06:00 AM

This happens whether or not the new one is stronger or weaker than the old one.

Does it mean that it's actually possible to refresh a stronger rupture with a weaker one? No more "A more powerful buff is already present"?

#40 Lumen222

Lumen222

    Von Kaiser

  • Members
  • 57 posts

Posted 29 August 2012 - 05:02 PM

Does it mean that it's actually possible to refresh a stronger rupture with a weaker one? No more "A more powerful buff is already present"?


Yes.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users