search

simulationcraft / simc

Simulationcraft engine/GUI
GNU General Public License v3.0
1.4k stars 692 forks source link

Arms Warrior: PTR Slam damage too low #457

Closed navv1234 closed 9 years ago

navv1234 commented 9 years ago

Originally reported on Google Code with ID 458

It appears PTR Slam damage is far too low (SimC 403-30).

Earlier today I was investigating why some of us in the Arms community were getting
optimal 4.0.6 results by heavily favoring HS over Slam in SimC action lists, despite
HS receiving large nerfs and Slam equally large buffs in the patch. I spent many, many
hours on the PTR boss dummy, and my anecdotal findings were that Slam was performing
significantly better in practice. Looking for a mathematical explanation for this,
I saw that the Slam numbers in Recount were quite a bit higher than those given in
SimC with the same warrior ported in and the same self-buffs. With an adequate sample
size, here were the numbers I was getting for each.

Actual PTR Slam: 13973 average hit, 32751 average crit, 45225 max crit
SimC PTR Slam: 10473 average hit, 24694 average crit, 35771 max crit

I'm not sure what is causing this. I have checked many times that the buffs and character
are identical and that SimC is set to PTR. I'm not sure if this issue pre-dates 403-30
or if the issue carries over to Fury. I will attach the profile I am using, which is
a customized 4.0.6 rotation and PTR premade character so that you may recreate these
results if necessary.

Reported by remix1@live.com on 2011-02-05 22:41:26

navv1234 commented 9 years ago 
When I run that profile through simc v30 with ptr=1 I get:

   slam                Count= 52.2| 7.16sec  DPE= 15714| 9%  DPET= 10407  DPR=1134.4
 pDPS=1818  Miss=0.00%  Hit=12606|10280|17467  Crit=29356|23356|39686|18.6%

which is different from both of your numbers.....

When you run in the gui, it generates two files:  simc_gui.simc and simc_report.html

Could you include both of those files in the issue?

Reported by natehieter on 2011-02-05 23:51:37

navv1234 commented 9 years ago 
I'll attach those, but do you have the same buffs selected? Those look very similar
to the numbers I get with full buffs and debuffs. As I mentioned I was testing only
with self-buffs (sunder, bshout, and blood frenzy/trauma).

Self-buffed: slam                Count= 52.7| 7.81sec  DPE= 12369| 9%  DPET=  8048
 DPR= 898.9  pDPS=1299  Miss=0.00%  Hit=10473|9041|15744  Crit=24684|20541|35771|13.3%
Full buffs/debuffs: slam                Count= 58.0| 7.16sec  DPE= 15420| 9%  DPET=
10211  DPR=1115.3  pDPS=1784  Miss=0.00%  Hit=12432|10018|17467  Crit=28837|22760|39686|18.2%

Also, I couldn't help but notice that the Slam numbers were very close to 30% apart,
which is how much LttS provides to Slam on the PTR. Looking through the warrior engine
file, I found some PTR LttS code under MS, OP, and Execute, but not Slam. Am I interpreting
this wrong or is that the cause?

Reported by remix1@live.com on 2011-02-06 00:40:18

navv1234 commented 9 years ago 
Went ahead and plugged in a talent build without any points in LttS and ran it.

slam                Count= 52.4| 7.82sec  DPE= 12376|10%  DPET=  8052  DPR= 901.2 
pDPS=1293  Miss=0.00%  Hit=10471|9041|15744  Crit=24724|20541|35771|13.4%

Compare to self-buff numbers posted above. So LttS just needs to effect Slam for PTR
version and it's fixed.

Reported by remix1@live.com on 2011-02-06 00:53:31

navv1234 commented 9 years ago 
Sweet!  You get my Favorite User of the Week award!  We have a bit of indigestion at
the moment (tons of changes) so I can't spin a release immediately.

I have the code-snapshot for v30.  I'll put out a v30.1 from there in a couple hours.

Reported by natehieter on 2011-02-06 01:04:08

navv1234 commented 9 years ago 
This issue was closed by revision r6755.

Reported by delmortis1 on 2011-02-06 01:24:08

navv1234 commented 9 years ago 
I've fixed this in r6755

Reported by delmortis1 on 2011-02-06 01:24:48