Mar 182011
 

Greetings to the feral community. I am going to keep my intro to a bear minimum (pun intended). I started playing WoW about 3 weeks after the game was released. My first level 60 was a rogue with about 40 days played. My second level 60 was a feral druid with over 378 days played, otherwise known as “Leafkiller.” Let’s just say I liked playing the druid more. Currently I maintain an Ovale script that is posted in the cat forums of The Fluid Druid and can be found here. My script is based on the work of many ferals over the years dating back to the early move suggesters such as FBN. In keeping with the tradition that was set by nightcrowler, my script has been tuned by using a cat DPS simulator (I use Mew). If you want more background on Mew, read the Mew N’ You post by Qbear.

For better or worse, my posts will generally be focused on answering questions through simulation testing and will be a bit on the dry side. It is in my nature to prefer hard data to opinions. Many times on the forums I see people post answers that lack a numerical foundation and more often then not, the answers are incomplete, wrong or in some cases about something that is trivial and not worth discussing.

A question I see asked a lot is what secondary stats people should reforge to. Before 4.06 the question was pretty easy to answer – mastery was significantly ahead of the rest of the secondary stats and for most people crit was the clear second choice. 4.06 changed that by making our DoTs weaker and Shred/Mangle stronger. Mastery has come down in value while stats such as Hit and Expertise have come up in value. Back in October, when 4.0 first came out, I posted on EJB about dropping hit and expertise from my gearing and seeing a dps increase of over 500 (this was at level 80). The ensuing discussion was pretty lively, and (to me) culminated in a post from tangedyn which he concluded with the following statement:

It’s time we drop the notion that it is necessary to cap hit and exp. The game has changed, and we either adapt or fall back in the damage meters.

With the changes in 4.06 bringing the secondary stats closer in value to one another, I decided to retest exactly how much the dps cost was for reforging to the hit cap. 500 dps at level 80 was (to me at least) pretty substantial. It was 3% of my total damage. With 4.06 I knew the numbers had changed substantially and I wanted to know what the cost was to reforge to the hit cap. There are advantages to being hit capped which include not missing interrupts (changing in 4.01), not missing cowers, and a more stable dps rotation. There have also been many posts where people have claimed better dps on add fights with more hit. I did not try to get both hit and expertise capped as that would have resulted in gemming/enchant changes. One thing that has not changed with 4.06 is that the relative value of agility is still around three times that of any of the secondary stats and the relative value of strength is still more than twice that of the secondary stats and there are not enough secondary stats available to reforge to get both hit and expertise capped.

To test the impact of emphasizing crit, haste or the hit cap, I reforged my current gear the following four ways:

  1. First I emphasized crit rating after mastery.
  2. Then I emphasized haste rating after mastery.
  3. The third setup had haste rating and crit rating almost equal.
  4. For the fourth setup, I got as close to the hit cap as I could (960 hit rating for 7.99% hit) which did result in a slightly lower mastery rating.

I tested this against four combat scenarios as follows:

  1. A 5 minute patchwork style fight (“Patch”)
  2. A 5 minute patchwork style fight with in combat Ravages with no travel time (“Ravage 0″)
  3. A 5 minute patchwork style fight with in combat Ravages with 2 second travel times (“Ravage 2″)
  4. A 7 minute fight that alternates 90 seconds of combat with 30 seconds of downtime which is similar to the air phases on Atramedes (“90/30″)

All tests used the Glyphs of Rip and Shred, and I did full runs with the third glyph being either TF, Berserk, or SR (this is captured as row prefixes in the data below). The testing I did is not intended to comment on the relative merits of the TF, Berserk, and SR glyphs, but was done to make sure that the behavior of a hit capped profile was consistent regardless of which of those three glyphs was used. I will probably post a followup to discuss the relative merits of each of those glyphs. I have included all of the data I measured at the end of this post. I did the testing with my Mew script, which is posted here. The differences between my script and the default one that ships with Mew are pretty minor with my script typically producing slightly higher dps. To emulate the Atramedes air phases I added some code to the top of the script to return a null result after every 90 seconds for 30 seconds. I made the fight length 7 minutes to ensure that the fight ended during a combat phase.

Looking at the data below, the differences in value between haste and crit are low (the largest difference was with the glyph of berserk on the 90/30 fight where emphasizing haste comes out about 50dps ahead of stacking crit). Reforging to the hit cap is a dps loss in most scenarios – but it is not nearly as significant as it was before the 4.06 patch came out. Using it in conjunction with the glyph of TF generally showed about a 60dps loss while using it with the glyph of berserk was closer to a 100dps loss. It would be a mistake to make too many assumptions about the glyph of berserk results as it is particularly sensitive to fight length and mechanics (for example on the 90/30 fight, Mew is showing less than 3 full berserks which means the second berserk is being clipped). What was interesting was that with the the glyph of TF on the 90/30 fight the hitcap setup was on par with the other setups – which lends some credence to the notion that being hitcapped helps with dps on fights with target switching/adds.

The bottom line is that the dps loss for reforging to the hitcap is about .25% of the measured dps. The secondary stats are so close in value now that you you can pretty much choose any combination of crit/haste/exp/hit without significantly impacting your dps. Mastery is still preferred, but even giving up a small amount of mastery to get hit capped is not that significant (I had to give up 91 mastery with my gear).

-Leaf

*Updated the data with agility at 5060 not 5360.

Crit Haste Balanced Hitcap (7.99%)
Weapon DPS 703.5 703.5 703.5 703.5
Str 142 142 142 142
Agi 5060 5060 5060 5060
AP 887 887 887 887
Haste 663 1260 1134 782
Hit 266 266 266 960
Crit 1604 1007 1133 957
Exp 188 188 188 113
Mastery 1882 1882 1882 1791
TF Patch 23562 23588 23580 23529
TF Ravage 0 24797 24813 24807 24759
TF Ravage 2 24154 24142 24152 24048
TF 90/30 20127 20165 20147 20170
Berserk Patch 23492 23527 23514 23418
Berserk Ravage 0 24593 24598 24601 24526
Berserk Ravage 2 23994 24008 23995 23883
Berserk 90/30 19859 19909 19870 19827
SR Patch 23429 23462 23459 23372
SR Ravage 0 24534 24547 24536 24469
SR Ravage 2 23937 23925 23917 23840
SR 90/30 19797 19842 19811 19783

  23 Responses to “4.06 – the Homogenization of Secondary Stats”

  1. Thanks for the data. It’s both good and bad to see the current state of secondary stats. Good because you can choose, bad because that choice will largely be irrelevant. From this point of view it’s a bit sad that Blizzard is moving away more and more from the “preparation” part of the fight (gearing, optimization) while going more and more towards the “execution” part of the fight.

  2. Thanks for all the testing Leafkiller :)

    I have a question though. Does it make sense to go for the hit cap without going for the expertise cap aswell? If its that important for a player not to miss , that he is willing to go down in dps (even though its not a lot), it would seem he hasnt really accomplished anything without going for the expertise cap aswell.

    In any case 4.1 will change all this, when we cant miss on the interrupt.

    • It takes 960 hit rating to hit the cap (8%) and 780 expertise rating to hit the dodge cap (6.5%). There are not enough secondary stat points to reforge to both the hit and expertise caps (at least on my gear). Trying to go for both would result in a much larger dps loss as you would have to start trading out agility and strength in addition to trading out mastery.

  3. Leafkiller, are you sure of your agility values? They seem a little high.

    • Shoot – you are right the agility is 300 too high. Clearly I still had a flask on from raiding the night before. The other numbers are all correct as I did actual reforging rather than try to work out the changes by hand.

      Fortunately that will not change the relative values enough to affect the outcome. It is pretty annoying to find that I missed that since I try to be very methodical when doing measurements. Clearly senility is upon me…

  4. I’ve always wanted to try this out, a while ago I ran into a forum thread on the US forums complaining not about stateofdps.com but actually this website http://www.simulationcraft.org/410/Raid_T11_372.html
    I knew our stat weights were close but it wasn’t till looking through simcraft that I noticed how close they really were. I guess I can’t complain to much at least our feral brothers stuck with kick duty in 10m can remain hit capped for all fights and not have to worry about a massive dps loss.

  5. Let me just say Leafkiller’s Ovale script is a must for new cats. It is one thing to explain the “rotation” and another to see it. Download it, install it, and practice it. But don’t just mash it. Pay attention to what it is doing and most importantly WHY it is doing it. This will make you a better cat. Thanks for all your work Leaf and welcome to the team.

  6. With 4.1 I’ll be going expertise capped and hit when I can. With interrupt not missing, I find expertise a lot better should I find myself DPSing from the front, which can happen at times. In that scenarion, expertise is worth double the hit (which puts it well ahead the rest).

    @Qbear: LOL at the “massive dps loss”. 5% is not massive by any stretch of imagination….

  7. I’ve been trolling TFD forums for a few months now, and I’m excited to see Leafkiller contributing to the actual blogs now as well.

    Looks like the team that took over is going to be nicely rounded.

    High end raiding, theory crafting, and class mechanics discussions for feral dps specifically has always been hard to come by in an openly accessible and public format.

    I’m looking forward to further posts about the state of our class at all tiers of raiding, and hoping that developers can see our complaints aren’t all random QQ, but are backed by data and experience with current content.

    Thank you Leafkiller for your post today, and I look forward to seeing more from you in the future! :)

  8. Thanks for all the work. I really liked reading this. It does point to one interesting conclusion. No longer is gemming/enchanting/reforging to the precise amount of crit/haste/hit matter. And with our ‘rotation’ pretty much ironed out and dumbed down this means the only way for ferals to really increase their DPS is for Blizz to make changes.

    I’ve noticed that the standard deviation for top ferals has really shrunk. Some of the difference would always be attributed to gear/gemming etc. But in Wrath a big portion was suspected to be due to how well you nailed the JFM rotation along with the RNG impact like OCC procs.

    Out standard deviation is now much closer together which means blizz has taken out things like rotation/gemming and even RNG impact is most likely reduced. So what does that mean?

    To me it means for us to get higher on the DPS charts we need for some changes to our spec. This type of work pretty much removes the issue on if we are really achieveing our maximum potential. The answer seems to be yes and unfortunately our maximum potential is not good enough compared to others.

    • I agree, the numbers do speak for themselves.

      I miss the days of min/maxing meaning something for feral dps as well as the “JFM” rotation separating a “good” feral from a “great” one.

  9. I’ve always felt that the value of hit/expertise has been undervalued and not properly represented in Sims. My reason behind this is that it’s not that missing is necessarily bad, it’s when you miss that can be terrible. I’m sure that’s why you see the numbers are so close when you’re switching targets.

    I remember back in ICC, I missed 3 mangles in a row at the beginning of the Blood Queen fight. That really hurt. Had I missed a mangle at any other time during that fight, it would not have mattered anywhere near as much.

    All that to say, I’m glad to see Leaf has shown the numbers to back up what I’ve felt for a while now. It means I can reforge back to hit without gimping my raid :)

    • I hear that, I was testing my dps on target dummy yesterday and missed my first 3 FFFs. If this had happened in a raid it would’ve hurt my dps and the tank’s threat quite a bit more than just missing a random shred mid-combat.

    • Actually, the place where hit/expertise are real good is in fights where you have to focus on other stuff, like movement or survival. Since you cannot keep your eyes fixed on timers and the CP number, some “flying blind” is required. When things go wrong is a moment where you find yourself at 4CP, a OOC proc and highish energy. The normal approach is a fast shred/rip, but if by any chance the shred misses, unless you keep checking your CPs you risk applying a 4 CP rip. And the interface has no feedback whatsoever indicating the rip CP application, the only way is to look at the inflicted damage and run some math in your head, which in a complex fight is a good way to get killed.
      A 4 CP rip is a lot of dps lost…. (something of the order of 3k per tick, without considering criticals), for comparison, even before 4.0.6, the difference between hit and crit/haste was around 0.2 DPS/point, which means 200 dps (1% of 20k) from going hit capped to 0 hit rating. 1% is less than the RNG deviation…. so unless your guild is downing bosses by the RNG, you were never gimping anyone’s raid by going full hit….

  10. Thanks for the great work, Leafkiller, and really nice to see you contributing to more articles.

    I also appreciate you talking about adds/switches and the value there :)

    I guess I’m kinda sad about the homogenization of all stats; it basically means that I shouldn’t bother looking at gear other than ilvl and sockets. That’s fairly boring to me. I miss the days where getting a specific piece really meant something; now, only weapons really do that. But I know I’m somewhat unusual in that regard.

  11. Leafkiller, your icon needs more leaf killing

  12. isn´t there something wrong with the mastery value as well?

    if you reforge to 960 hit rating, you should have lose much more mastery rating: 1882 –> 1579

    are your calculations still correct then?

Leave a Reply