Sep 142009

This post is a continuation of part 1, where I explored whether a feral DPS character should gem/eat for Agi/ArPen. The conclusion from that post was that agility was a superior choice; left open, however, was the question as to whether better supporting gear would make the switch to ArPen more viable. I’ll look at that today, briefly discuss my thoughts about the playstyle differences between Agi and ArPen, and make some final conclusions.

Agi vs. ArPen…Round 2

For this test, I went to wow-heroes and whistled up some much better geared ferals than myself. We’ll compare Numa, Wildoak (both from Magtheridon) and myself, in my updated gear. See table below. For a better comparison, I’ve subtracted 200 Agi/200AP and added 200 ArPen from my Armory stats in order to simulate having gemmed ArPen as well.

Character: Numa Wildoak Alaron
Wow-Heroes Score: 2880 2604 2434
Attack Power: 8749 8464 7989
Agility: 1932 1830 1641
Strength: 114 112 167
Crit Rating: 846 476 717
Hit Rating: 233 260 211
Expertise Rating: 130 105 42
Haste Rating: 509 203 190
ArPen Rating: 530 641 455
Avoided Attacks: 0.93% 0.87% 4.28%

We’ll plug each kitty’s stats into Toskk’s Cat DPS calculator, and evaluate whether +1 ArPen vs. +1 Agi results in more DPS, under the following 3 cases:

  • Case 1- 10man buffs: Flask of E. Rage, F. Feast, Imp. MOTW, Horn of Winter, Heroic Presence, BoKings, Sunderx5, FF
  • Case 2- 25man buffs: add Imp. BoMight, Windfury, Unl. Rage, Bloodlust, F. Inspiration, Imp. Moonkin Form, Heart of the Crusader, Mangle/Trauma always up
  • Case 3- 25man buffs, with Grim Toll or Mjolnir Runestone equipped

Admin notes: All cats assumed to have all DPS talents, Relentless Earthsiege meta, Mongoose enchant, Rip/Shred/SR glyphs, and are given equipped idols/ set bonuses/trinkets. For Case 1 and 2, cats with ArPen trinket (Numa/WO) given Mirror of Truth proc instead. For case 3, Alaron/WO given GT proc, Numa regains MR proc. All settings default, except all OoC procs for Shred checked.


  Case 1 Case 2  Case 3
Numa 0.93 1.02 1.09
Wildoak 0.89 0.98 1.03
Alaron 0.82 0.89 0.93


And here’s the relevant DPS numbers, just for grins:

  Case 1 Case 2  Case 3
Numa 6868 9092 9236
Wildoak 6158 8127 8237
Alaron 5771 7501 7801



As you can see above, whether to switch from gemming Agi to ArPen depends on three factors; your gear, your raid, and an ArPen trinket.  You can’t simplify the answer to just “if your ArPen > X, switch.” For example, Numa has good enough gear to where he no longer needs the trinket for ArPen gemming to be an upgrade, whereas Wildoak is on the bubble, where either Agi or ArPen could make sense, even though Wildoak has more ArPen then Numa. This is because ArPen’s value also scales better with other stats, mostly AP, hit, and expertise, than Agi. As for me? I’m not even close to considering a switch.

Final thoughts:

ArPen has some very important things going for it. It’s more consistent then Agi; you get less crits, but they all hit harder. This is good for encounters with tight DPS requirements, where a string of “good” luck can pull aggro, or burst DPS, where Berserk/Heroism/trinket CD’s generally favor ArPen to Agi. Of course, the converse applies on other fights, where sustained DPS/high-movement DPS dilutes CD effectiveness. You also give up some emergency off-tanking viability, and the reduced crit makes maintaining uptimes on all your effects slightly more challenging.

Lastly, put it in context. For all the discussion about it, we’re talking, at best, a difference of +/- 300 between Agi and Arpen. Unlike 3.2.1 where some well-geared cats were seeing a 25-30% difference in gemming ArPen for Agi, it’s a much less profound difference now, and you’ll only see a DPS difference of 3-5%, at best (and that’s well within average DPS variability, so you may not notice an increase at all).

Overall, ArPen remains a good stat for geared ferals, but I wouldn’t recommend gemming for it on a general basis. Obviously, if  you have absolutely insane gear like Numa above, it’s a much tougher choice, but I think I’d still prefer agility.  Don’t take my word for it; plug your stats and usual buffs into the Calculator and decide for yourself, based on your situation. Assuming the 3.2.2 numbers remain the same into Icecrown, however, the better gear available will tilt the pendulum back towards armor penetration again…especially if ferals start softcapping agility.

Sep 102009

EDIT:  See Part 2 for a more in-depth discussion.

Q: WTF is up with this armor penetration stat?

A: Most of your character’s attacks are affected by your target’s armor. Armor Penetration enables these attacks to ignore a certain percentage of armor.

Q: So, like, which attacks?

A: Generally, direct attacks are reduced by armor (and thus increased by ArPen), while DoT effects, such as kitty bleeds, are not.

Q: That doesn’t seem too complicated. Why all the bitching?

A: Because it scales in value dynamically, compared to other stats which scale in value linearly.

Q: Right…what?

A: Ok, let’s explore some other stats, then come back to ArPen. Bring in Example Kitty! Here are Example Kitty’s unbuffed stats, which are based on a character snapshot I took in late 3.1 (FWIW, Naxx-25/Uld-10/Uld-25 gear, no hardmodes), and fed into Toskk’s Cat DPS Calculator. (Available at work!)

  • Str: 248
  • Agi: 1355
  • Hit: 236
  • Haste: 226
  • AP: 7674
  • CritRating: 702 (54.03%)
  • Expertise: 38
  • ArPen: 204
  • Talents: All
  • Glyphs: Shred/Rip/SR
  • Buffs/Debuffs: None
  • Bonuses: Relentless Earthsiege meta, Mongoose chant, Idol of the Corruptor, 2pc T7, DM:G trinket, Mirror trinket.
    Settings: Default, except all OoC procs for Shred checked.

According to Toskk’s Cat Calculator, I should be able to pull 4407 DPS with these stats. Let’s look at AP, to start. AP scales linearly, so here, with every point of AP I add, I gain 0.363 DPS. No matter how much AP I gain (or lose), if all my other stats remain the same, each point of AP is worth the same amount.

For something a little harder, let’s look at Hit. Hit scales linearly as well, but has a cap. Currently, 1 point of hit is worth 0.597 DPS to EK, and that stays the same, regardless of whether i have 0 hit or 261 hit. Above 261, (or to be exact, 262.3), hit is worthless, since you can no longer miss a raid boss.

Now, finally, ArPen. ArPen works much differently, in that the value of individual points starts out very small, but increases as you accumulate more, until it finally hits a cap of its own. This cap is currently at 1230, and will be increasing to 1398 in 3.2.2 (more on this in a sec).

So, keeping EK’s other stats current (3.2.1/3.2.2 values):
With 0 ArPen, 1 point is worth 0.57/0.50 DPS;
With 300 ArPen, 1 point is worth 0.67/0.58 DPS;
With 600 ArPen, 1 point is worth 0.81/0.68 DPS;
With 900 ArPen, 1 point is worth 0.99/0.81 DPS;
With 1200 ArPen, 1 point is worth 1.25/0.98 DPS.

Remember that this is unbuffed; buffs/ArPen trinkets/debuffs can change the math significantly (see next Q)

Q: And these numbers are important because…

A: Well, they’re only relevant in context with Agility. Normally, we want to stack +agi over everything else, as +agi gives us our biggest boost per point. However, at a certain ArPen/stats level, +ArPen becomes more valuable. Let’s run a few more examples:

Case 1: EK, unbuffed as before.
Case 2: EK, 10-man buffs/debuffs (Flask of E. Rage, F. Feast, Imp. MOTW, Horn of Winter, Heroic Presence, BoKings, Sunderx5, FF)
Case 3: EK, 25-man buffs/debuffs
(Same as Case 2, add Imp. BoMight, Windfury, Unl. Rage, Bloodlust, F. Inspiration, Imp. Moonkin Form, Heart of the Crusader)
Case 4: EK, 25-man buffs/debuffs, swapping in Grim Toll for Mirror of Truth.

 The following data tables/graphs are measuring the effectiveness of stacking +ArPen over +agi, measured by percent effective. (For instance, a result of .75 means that adding ArPen, in this particular case, has 75% the effectiveness of adding agi.)

Currently (3.2.1):

ArPen 3.2.1 Case 1 Case 2 Case 3 Case 4
0 0.74 0.77 0.83 0.89
300 0.83 0.86 0.94 1
600 0.94 0.97 1.07 1.14
900 1.08 1.11 1.21 0.93
1200 1.24 1.28 1.39 1.39


Upcoming (3.2.2):

ArPen 3.2.2 Case 1 Case 2 Case 3 Case 4
0 0.64 0.67 0.74 0.77
300 0.72 0.75 0.82 0.86
600 0.8 0.83 0.91 0.96
900 0.9 0.93 1.02 0.78
1200 1.03 1.05 1.15 0.89

Arpen3.2.12 width=

As you can see, 3.2.2 really kills off the idea of ArPen stacking for lower gear levels, with your average 25-m raid requiring ~900 ArP from gear in order for stacking it to be an improvement. Remember, all of this will scale with gear, so feel free to enter your own numbers into the calculator.

Q: TL;DR. Sum it up for me!

A: Sigh. Okay. Assuming you have ~ivl 215 gear, as it stands now (3.2.1), you want to switch to gemming/flasking/eating for ArP around 450 for 25m raids, or 300 if you have Grim Toll/Mjolnir’s Runestone. (high 600’s for 10m raids.)
For 3.2.2, you would switch at 850ish for 25m raids, and 1150ish for 10m. For GT/MR, I need to test it some more and get more data points, but at this gear level, the proc is tough to manage since you have to (attempt to) stack so much ArPen for it to be useful, but then much of the proc effect is wasted due to the cap.

In short: gem/flask/eat for agility.

EDIT: See Part 2 for a more advanced discussion at higher gear levels…the decision scales not only off your personal ArPen, but your other stats as well.

Sep 062009

From MMO-Champion:

The nerf to armor pen in 3.2.2 is intentional. Compared to the recent buff where we increased the value of armor rating to 125%, this nerf would take it back down to 110%. While we are still evaluating the effects of this change in the 3.2.2 build, we did want to let you know of the possibility in case you were about to spend a lot on armor pen gems.

In fact, this was really the point. Several melee specs (and Marksman hunters) had begun to focus on armor pen at the expense of all other stats. Gear without armor pen was being passed over and gem sockets were increasingly being filled with just this one stat. While every spec has stats that are more valuable than others, this one felt like it was starting to trump everything. Not coincidentally, characters stacking lots of armor pen were starting to do more damage than their peers and more damage than we were comfortable with.

This change is largely for PvE reasons, though we won’t cry at all if melee damage in PvP drops a little as a result.

We’re letting you know now so that this doesn’t feel like a stealth nerf, assuming it goes live. While you might disagree or be frustrated by the change (though I also suspect it won’t come as a surprise to many players), we ask that you try and keep your response to something appropriate for these forums.

Assuming this stays, this is a 12% nerf to ArPen. It’s worded a bit oddly…essentially they’re rolling back the previous 25% buff to make it only 10%.

What does this mean? I’m sure Kalon and Runy will weigh in, but as it stands now, ArPen was insane for geared melee classes, because it scaled up rather than scaled down. Basically, every point of ArP was worth more than the point before, so once you start getting high levels of ArP, adding more ArP was giving HUGE dps gains. (Sims were showing cats in full BIS gear at 10k DPS, since they were able to hardcap ArP.) The theorycrafters are still checking, but I’m almost positive that stacking agility will return to being the best strat for almost all gear levels.

Sep 042009

Mitigation and Avoidance

Ah, the most familiar leg of the triangle. Mitigation and avoidance, combined with healing (see Part 1), is how you survive long enough to actually take down boss massive HP pools solo, since your healing in feral form isn’t that great. As a bear, you have three main sources of M&A: Armor, Savage Defense, and Dodge.


  • Example Boss hits for 10k damage per attack, and attacks every 2s, giving him 5000 DPS (unmitigated).
  • We have 40k HP, 30k armor, 5000 AP, 42% dodge, and 40% crit.


Armor (or Damage Reduction) is your most important defensive statistic. Simply put, it reduces the damage from incoming melee attacks, which is typically the largest damage component of most fights. The formula for Armor is Armor*100/[Armor+(467.5*AttackerLevel – 22,167.5)], which can be reduced to [Armor / (Armor + 16,635)] x 100 for boss level mobs (skull = your level +3). Armor also has no diminishing returns…while it appears to (since you need more and more armor for each 1% of DR), each 1% of DR is more valuable than the last. Let’s use an example:

Using the assumptions above, with 0 armor, you die in 8 seconds. With 15k armor, you have 47.4% DR. This reduces EB’s DPS to 2629 and kills you in 15.21 seconds, for a 7.21 sec increase. With 30k armor, you have 64.3% DR, which reduces EB’s DPS to 1784 and kills you in 22.42 seconds, for a 7.21 sec increase. Got it?

Remember: Armor is useless against magical attacks…and bosses that debuff your armor (like Prince Malchezzar) can suck.

Savage Defense

Savage Defense is a very interesting mechanic. Basically, every attack that crits (including bleeds with Primal Gore) will proc SD, which will reduce the next hit by 1/4 of your attack power. Your basic rotation gives you 1.3 times to crit per second (counting melee/Maul, Mangle/Lacerate/Swipe, and Lacerate bleed, not acounting for haste). This means that in general, assuming you can keep up a full DPS rotation, you should have around 75% of hits affected by SD on a single target, and more on multiple targets. (Kalon at Thinktank has done much more extensive theorycrafting on this.) If you have 5k attack power, that means 75% of enemy attacks will be reduced by 1250, after they are already reduced by armor.
Essentially, what this means is that SD is GREAT for low-level stuff, and meh for raid-tanking. Continuing our example above, the boss currently doing 1784 DPS will now do 1315 DPS, which is roughly an extra 25% mitigation. Against something that hits really hard (like Patchwerk, for example, who does like 80k damage per attack), a 1250 damage reduction is a drop in the bucket, and gives you MAYBE 0.5% of additional mitigation.
Also, getting stunned sucks. I’ve actually died to Moroes trash when I got stunned and hit with the 0 armor debuff at the same time…ouch.


Avoidance is pretty simple for druids. Each boss melee attack has a 7.8% chance to miss + your dodge, which is usually slightly above 40%. I won’t do a full-on dodge analysis…just remember that it comes primarily from your agility, plus any extra dodge rating youy have.  Not counting agility procs (Mongoose, DM:G), you can generally expect about 50% avoidance while soloing, and much more on trash (which will have a much higher miss chance). Assuming 50% avoidance neatly reduces EB’s DPS in half, to 657 DPS.

Indestructible Potions

Indestructible Potions are GREAT. They give you an extra 3500 armor, and can be used twice (if you use one pre-pull). Using the stats above, an extra 3500 armor should reduce EB’s DPS to 595, which translates to 7440 HP saved, or twice that with two. A great improvement over a healing potion. They’re not as great for raids, where the threat is typically burst, not sustained damage, but are still very helpful.


Using the Healing calculation from part 1 (336 HPS), EB only does 321 sustained DPS, which gives you over two minutes to kill the boss. That’s where DPS stats comes into play, in the next installment.

Aug 262009

I’m working on a long guide talking about soloing content as a Druid. Basically, I’m trying to model each leg of the “holy trinity.” If we know our DPS and boss HP we can estimate boss TTL (time to live). If we know the bosses’ abilities and our mitigation/avoidance/resistance we can model DTPS (damage taken per second), and with calculating our HPS and max HP, model our TTL (time to live). If  our TTL > boss TTL, we win. Obviously, skill/RNG/lag/whatever will cause deviations from the math…but that’s okay, WoW’s still a game, not a spreadsheet. :) This is not intended to be EJ-level stuff, BTW, this is just me having fun with math and trying to learn a little more about WoW mechanics in the process. In any event, here’s the start of the HPS section.


  • 40000 hp (not unreasonable solo, with MOTW/stoneblood/food buff/stamina scroll/profession bonii (AND new LW kings drums…woot patch notes)
  • 40% crit
  • Unlimited rage (boss tanking)

Improved Leader of the Pack

Improved Leader of the Pack is your primary self-heal, and is critical for boss soloing.  It is affected by Nurturing Instinct, but only in cat form. I’m not positive, but I’m pretty sure it’s affected by Frenzied Regeneration when glyphed.

  • 60 attacks per minute capable of proccing ILOTP (20 Maul, 40 swipe/lacerate/mangle/FF…sadly, Lacerate bleed crits from Primal Gore do not proc ILOTP, but they do proc Savage Defense.)
  • ILOTP has 6s cooldown; procs on average every 7.5s or 8 attacks per minute. This is probably a bit conservative of an estimate (doesn’t account for haste, math’s a bit rounded) but it’s close enough.
  • Therefore, 40k hp = +1600 HP ILOTP tick * 8 = 12800 hp = 213 HPS. If you assume 1.33 ticks (on average) are buffed by glyphed FR and glyphed SI (see below), that increases to about 220 HPS.

Frenzied Regeneration/Survival Instincts

Frenzied Regeneration is your secondary self-heal, and is key to your tactics. Pretty much, for most bosses, if you can live long enough to Regen twice, you’ve probably won.  Normally, SI by itself is fairly useless when soloing. However, when combined with FR, it makes your FR tick for more (since it uses the new higher HP value for calculating healing done). Occasionally, for dps race type bosses, SI may be better used as a last-ditch spurt of extra time to kill the boss. I’ve calculated the healing done from all the abilities singly, in combination, and with the glyphs: see below.

Abilities Max HP Total Heal (HP3M) HPS
FR 40000 12000 67
FR (Glyphed) 40000 14400 80
FR + SI 52000 15600 87
FR + SI (SI Glyphed) 58000 17400 97
FR + SI (FR Glyphed) 52000 18720 104
FR + SI (Both Glyphed) 58000 20880 116

FR is affected by spell crit, and ILOTP is buffed by the FR glyph, so these numbers are a bit low.

Healing Spells

In theory, every boss should be easy- when you get low, just heal yourself, right? Sadly, there’s a couple big problems. First, your armor turns to tissue paper in caster form. Second, you have no healing talents, and no healing gear (unless you talent Nurturing Instincts). Third, you lose all your rage. No rage = no attacks = no Savage Defense (bad.)

With that in mind, many bosses have pauses where you can safely get a cast off and shift back to bear.  In general, if it’s a pause where you can stay still, Regrowth is your best bet, if it’s a moving pause, Rejuvenation. For a long pause, Tranquility is great.  For example, against Moroes, once you kill the adds, he’s quite easy if you shift out to drop a Regrowth every time he vanishes, which will nicely help counteract the bleed damage. 

Now, if it’s a boss who attacks with primarily magic attacks, then your armor’s useless anyway. In that case, healing spells + NI become a lot more vital.  I haven’t had a chance to try it, but I’d imagine you could macro a swap to a healing weapon to get extra spellpower, Regrowth+Rejuv, and go back to cat (which, if you have NI, buffs HOT’s by 20%).

Other Healing Stuff

Lifeward is a good situational enchant for trash, but doesn’t proc enough on a single target to be worth it on bosses. Blood Draining is good in theory, but the threshold to trigger it is too low for you to make optimum use of it with Frenzied Regeneration. I’d stick with the +agi from Mongoose. Likewise, it’s better to use Indestructible Potions for most fights, rather than health potions. Pot right before the pull then pot during fight gives you almost 4 minutes of extra armor.


Currently, though, we know that if we have glyphed SI and FR (and there’s not much other choice for bears, really) the boss must break 336 DPS (shown in Recount as DTPS) to overcome our healing ability. If the boss does 436 DPS, for example, then we have a -100 HPS delta, which means we have a TTL of 400 seconds, and must then kill the boss before that…and that’s where DPS stats come into play. Stay tuned for Part 2, where I break down mitigation statistics.