A Call to Arms - Cataclysm Mechanics testing

[Redleg]

I was expecting the 83 (raid) dummy actually, but it only affects physical damage (so HotR base and melee). And I can correct for it in the code anyhow, so now that you told me I can go back and re-check it.

Grrr, sorry about that! Guess that will teach me about skimming through the posts.

[Warsadin]

Well at least he can edit the code to reflect the differences in dummies. Can't wait to see how it turns out.

[Klaudandus]

Well, my parses are gonna have to be from the lvl 80 dummy then, cannot find a heroic dummy free for me to hit

Lvl 80 BE
0/34/0 Spec
Dalaran Sword and white shield (no str, just armor) - No gear otherwise.
Was not trying to get top uptime on all my abilities, wanted to make sure I met certain criteria.

http://www.worldoflogs.com/reports/nsgncab5aclwd66m/ - Always Crusader Strike, Always ShoR with just one HoPow
http://www.worldoflogs.com/reports/zir5et0zp5peeb6y/ - Always HotR, Always ShoR with 2 HoPow

Next. Whatever I can hit but always 3 HoPow for ShoR.

[steadypal]

i wonder why they would nerf as dmg if 3sec cs/hotr arent even on beta yet

[Arcand]

http://www.29tharmor.com/resources/PTR-Oct-08-01.txt
http://www.29tharmor.com/resources/PTR-Oct-08-02.txt

First one is 2890AP, 889SP, 532-705 weapon damage (178.9x1.60 DPS - weapon is Shiver).
Second is 3179AP, 889SP, 565-738 damage (same weapon). For both of these I had 2/2 Seals of the Pure, 2/2 Wrath of the Lightbringer and 3/3 Crusade. Level 80 vs. level 80 dummy on PTR. No damage-affecting debuffs.

I screwed up several times and popped a sub-3 Shield. Sorry about that. If these are unusably screwed up let me know and I'll do another batch.

[theckhd]

Redleg's first data set:

Code: Select all

ShoR     626   662     36     6
CS       176   175     -1    -1
JoT      272   291     19     7
AS      3049  3169    120     4
HW      2354  2351     -3    -0
HoW     1288     0  -1288  -100
Exor    1155     0  -1155  -100
SoT       20    21      1     7
SoR       15    16      1     7
SoJ        7     7     -0    -1
Cens     430   460     30     7
Cons    1286  1337     51     4
HotR      21    21     -0    -1
HaNova   717   799     82    11
Melee     69    69      0     0

Redleg's second data set

Code: Select all

ShoR     972   978      6     1
CS       234   236      2     1
JoT      414   412     -2    -0
AS      3190  3160    -30    -1
HW      2381  2212   -169    -7
HoW     1344     0  -1344  -100
Exor    1199     0  -1199  -100
SoT       25    25      0     0
SoR       23    25      2     7
SoJ       11    11     -0    -4
Cens     663   655     -8    -1
Cons    1428  1430      2     0
HotR      26    26     -0    -1
HaNova   748   791     43     6
Melee     86    86      0     0

towellie's data set:

Code: Select all

ShoR    4416     0  -4416  -100
CS      1190     0  -1190  -100
JoT     1835  1835      0     0
AS      4585  4606     21     0
HW      2640     0  -2640  -100
HoW     1904     0  -1904  -100
Exor    1629     0  -1629  -100
SoT      144   145      1     1
SoR      117     0   -117  -100
SoJ       53     0    -53  -100
Cens    2987  2990      3     0
Cons    2842     0  -2842  -100
HotR     147   179     32    22
HaNova  1057  1318    261    25
Melee    492   496      4     1


Something weird is going on here, the melee and SoT damage is just about perfect,HotR damage is higher than they should be. All three are directly based on weapon damage, so that doesn't make sense.

Guess: Towellie was wearing 2pT10 and getting 20% extra HotR damage? That would explain the melee damage difference.

[theckhd]

Argh, after contemplating the weirdness of some of the results, I took a closer look and noticed that glyphs were still active. That explains the differences we were seeing in CS, Judgement, and AS. I'm going back through the parses from the last few days and correcting them now.

Klaud's first and second data sets:

Code: Select all

ShoR     626   625     -1    -0
CS       176   174     -2    -1
JoT      272   270     -2    -1
AS      3049  3038    -11    -0
HW      2354  2315    -39    -2
HoW     1288     0  -1288  -100
Exor    1155     0  -1155  -100
SoT       20    20      0     1
SoR       15     0    -15  -100
SoJ        7     0     -7  -100
Cens     430   426     -4    -1
Cons    1286  1284     -2    -0
HotR      21     0    -21  -100
HaNova   717     0   -717  -100
Melee     69    68     -1    -1

Code: Select all

ShoR     626   652     26     4
CS       176     0   -176  -100
JoT      272   268     -4    -1
AS      3049  3031    -18    -1
HW      2354  2316    -38    -2
HoW     1288     0  -1288  -100
Exor    1155     0  -1155  -100
SoT       20    20     -0    -1
SoR       15     0    -15  -100
SoJ        7     0     -7  -100
Cens     430   425     -6    -1
Cons    1286     0  -1286  -100
HotR      21    21     -0    -2
HaNova   717   753     36     5
Melee     69    69     -0    -0

This is using Human base stats, because I don't have the updated level 80 base Draenei or BE stats. In fact, I'd like to double-check all of the racial base stats (I suspect the STA value I have for humans is wrong), so if you have a level 80 paladin of any race on the PTR, please post the naked STR/STA/AGI/INT/SPI values.

[Klaudandus]

I dont use glyphs nor buffs. I shall see if I proc a faction change and use dwarf/draenei and repeat the experiment.

[Kierly]

Finally got my pally to show up on the ptr.

Here are the Human 80 naked stats. With 0/0/0 talents:

Code: Select all

Str:   151
Agi:    90
Sta:   143
Int:    98
Spi:   108


[Arianne]

Dwarf nekkid:
156 Str
86 Agi
165 Sta (Protection spec, so 15% sta inc) - 144 stamina without any spec
97 Int
104 Spirit


Draenei nekkid:
152 Str
87 Agi
164 Sta (Protection spec) - 143 stamina without any spec
98 Int
107 Spirit

[theckhd]

Arcand's first data set:

Code: Select all

ShoR    3468  3112   -356   -10
CS      1191  1075   -116   -10
JoT     1443  1443     -0    -0
AS      4200  4194     -6    -0
HW      1034  1072     38     4
HoW     1750     0  -1750  -100
Exor    1511     0  -1511  -100
SoT      111   110     -1    -1
SoR       92     0    -92  -100
SoJ       42     0    -42  -100
Cens    2347  2328    -19    -1
Cons    2452     0  -2452  -100
HotR     147     0   -147  -100
HaNova  1262     0  -1262  -100
Melee    376   382      6     2

Something's fishy here too. Melee, SoT, Censure, Judgement, AS all look fine. ShoR is probably just low because at least one out of the eight samples are not 3HP. CS is uncharacteristically low though. Are you sure you didn't have 2/3 Crusade? If not, I can't figure out why the model is giving an over-estimate

I had to tweak the Holy Wrath numbers a bit - one of the dummies you were hitting had a debuff:

Code: Select all

[21:36:22.304] Arcand Holy Wrath Grandmaster's Training Dummy 1 (O: 1072)
[21:36:22.304] Arcand Holy Wrath Master's Training Dummy 1 (O: 1073)
[21:36:22.304] Arcand Holy Wrath Grandmaster's Training Dummy 1 (O: 1159)

Given this, I've used 1072 as the HoW damage value and edited the model to pretend there were three targets. That seems to give good agreement.


Arcand's second data set - for this I've tweaked the Holy Wrath base damage to fit the data (2420 base @ 80):

Code: Select all

ShoR    3815  3821      6     0
CS      1267  1165   -103    -8
JoT     1540  1540      0     0
AS      4322  4305    -18    -0
HW      1073  1073      0     0
HoW     1793     0  -1793  -100
Exor    1554     0  -1554  -100
SoT      116   115     -1    -1
SoR       98     0    -98  -100
SoJ       44     0    -44  -100
Cens    2549  2496    -53    -2
Cons    2561     0  -2561  -100
HotR     155   151     -4    -3
HaNova  1262  1211    -51    -4
Melee    397   395     -2    -0

Crusade is still low, dropping to 2/3 Crusade fixes it but then makes HotR too low. There's clearly something going on with CS in Arcand's parse, or else my AP scaling for CS is doing something funny. I'm using normalized weapon damage (gear.avgdmg+player.ap*2.4/14), and since that's the only one of our abilities that scales with normalized damage it's possible something's wrong in there.

[theckhd]

Conclusions so far:
ShoR, Judgement, AS, SoT, Censure, HotR (physical) are all unchanged.

CS is likely unchanged, but we need to double-check the AP scaling. It could have been changed to unnormalized weapon strike.

The change to Holy Wrath's base damage doesn't seem to match up with Klaud's data set or Redleg's data set. I need to look at that more closely.

Hammer's holy portion has obviously changed, but we knew that already. I should have enough data now to work out what the new formula is.


New requests:

Test #1: Holy Wrath
This one is easy, since the damage is a fixed amount. Find a dummy with no debuffs, and record Holy Wrath damage for various spellpower levels. I.e. naked, ~500 sp, ~1000 sp, ~1500 sp, etc. Gear is basically irrelevant, since all we need is SP and HW damage. Since HW hits for a fixed amount, you don't even need more than one data point for each SP value, so this can be done with a pencil and paper instead of WoL. Make sure you note your spec - 0/38/0 preferred, but I'll take what I can get.

Then repeat the test from a spot where you can hit 3 dummies, again, no debuffs.

The level of the dummy should be irrelevant, but post it anyway just to be safe. I.e.:

Code: Select all

Source: lvl 80 BE
Dummy level: 80
SP      DMG (1 dummy)   DMG (3 dummies)
170     1000            350
501     2000            720
1000    3000           1100

Test #2: Crusader Strike
Setup: lvl 80 dummy is fine, Dalaran Sword for consistency. Record CS damage in various AP/SP configs, recording AP and SP for each. You'll probably want 20-30 CS casts per gear set. 0/38/0 build preferred. Probably easiest to log each trial separately and comment it with AP/SP (or post a link along with the details).

Example:
1000 AP / 500 SP : link
2000 AP / 1000 SP: link
2000 AP / 1500 SP: link
4000 AP / 1500 SP: link

Gear is pretty much irrelevant as long as you keep the weapon the same (Dal sword preferred just because I have it coded already - if you use something else just tell me what you used so I can code in the base speed and damage). Feel free to use Might or whatever other self-buffs you want to get different AP levels.

Test #3: HotR Nova

Same as #2, but for HotR. I can probably nail down the scaling from the data I have if nobody gets around to this, but one self-consistent set with the same weapon would make it a lot easier.

[towelliee]

Guess: Towellie was wearing 2pT10 and getting 20% extra HotR damage? That would explain the melee damage difference.

Nope and I wasn't glyphed for it either.

[towelliee]

Theck I also been uploading my recent heroics check them out I'd like to see inquisition numbers etc.

[angelflavor]

More Instance testing

http://www.worldoflogs.com/reports/teqjar6uzpn8qilz/

Last 2 I posted I was using Ret aura, which after looking at it only is like 1% of total damage... And thats with it dealing damage even when you block now.