[MoP] Beastmastery 5.2 - Page 4

Azortharion · 09/11/12, 8:16 AM

Originally Posted by Nooska

Having just completed a couple of tests on the "internal test bosses" where I let the pet attack the boss till focus was the limiting factor of basic attacks, then observing it for a while, then using RF, I feel confident that RF does affect pet focus regen (despite being ranged haste).
Specifically I observed mostly 5, sometimes 4 seconds between basic attacks (claw in this case) after letting my pet drain itself of focus. Popping RF increased the frequency to just over 3 seconds on average while RF was up. At no time did I myself attack the boss, so GftT was not a factor.

One thing I can certainly say as a biproduct of this testing, is that crit is worth a lot more than mastery untill we get enough GftT procs that the pet is never focus starved for basic attacks, this is likely the reason for thej variation in crit value that I observed at earlier simming.
That the pet doe snot generate enough focus to use basic attacks on cd also explains why haste at lower gear levels seem more attractive (valuation wise), because it affects pet focus regen twice - base focus regen and autoshot speed - neither one nor both combined will analytically make haste = crit at any point for pet focus though.

Will this mean that we will get to experience "crit soft caps" and such for MoP. E.g a specific point / rating that will (in theory) make your auto-shots crit enough to trigger GfTF enough to eliminate all delays in pet basic attacks because of focus starvation?

Naturally, RNG is a very determinate factor in this, and I suppose such a cap might be variable depending on lots of factors such as raid buffs.
Entirely hypothetically, however, what would this rating / percentage crit be to (again, in theory) let your pet never focus starve for basic attacks assuming that your auto-shot will -always- crit for the stated percentage of the time?

Nooska · 09/11/12, 9:42 AM

In reality there will be no softcap on crit, because whether you crit is rng, and therefore its is not improbable to not crit several autoshots in a row. All this means is that crit may be more valuable than shown in the sim if you are at high levels of crit, and be less valuable than shown when at a low level of crit.
We can do an averaging out and that way determine when you will on average never delay your pets basic attacks - this would be based on focus regen. Doing an averaging will make haste seem more valuable than it is though, since focus regen for the pet doubledips on haste - both from haste directly affecting focus regen, and from haste increasing the frequency of autoshots, thus increasing the frequency of critical autoshots and GftT procs as well.

Completely basic, to not delay pet basic attacks, you need a regen of 30 focus per 3 seconds, so 10 fps * fightlength (in seconds) - 120 (as you start with a full focus bar) for a 5 minute fight you need 2880 focus, or 9.6 fps.
At 0 haste your pet regenerates 5 fps, so we are left with 4.6 fps to gain from GftT, that equals 0.92 GftT procs per 3 seconds. Assuming a weaponspeed of 3 seconds we hit the GftT crit cap for 0 haste at 92% effective crit.

At 20% haste, our pet regenerates 6 fps, and out autoattack speed is at 2.5 second frequency. At 20% haste GftT needs to generate 3.6 fps for the same 5 minute fight, or on average 0.72 Autoshot crits per 3 seconds, which translates into a crit percent of 60% effective crit (1.2 autoshots per 3 seconds).
This means that 20% haste, in essence, generates the same amount of focus as 32% crit at those levels.

Summa summarum, we don't need to worry about getting to any crit softcap making it useless for GftT, even on average. This also shows why fervor is actually quite good for BM, and why we will want to use it essentially on CD, as it averages out to 1.67 fps for the pet, excluding the 50 focus gain - or the same as 0.334 GftT procs per 3 seconds, or 33% crit at 0 haste. (Remember, this is ignoring the instant 50 focus from using fervor)

NoGoal · 09/11/12, 10:47 AM

Even if pets could generate enough focus to never delay a basic attack, they still have Wild Hunt (2x focus cost, 2x damage) to make GftT still useful past that point.

Nooska · 09/11/12, 11:22 AM

I'm actually not that big of a fan of the current wild hunt, seeing as pets don't have focus to attack on CD, WH actually reduces the chance for Frenzy / Focus Fire uptime, because it eats (and damages) focus for 2 attacks, but only provides 1 stack of Frenzy if procced (currently), thus being a worse option than simply overflowing focus a bit. I have a hunch (need to check either simmed or analytically) that WH actually also devalues Fervor, due to the same mechanic - it translates directly to damage via WH, but Frenzy uptime probably suffers due to more WH uses during fervor - though the ticking extra fps may make up for it with keeping focus high enough to let basic attacks be cast on CD, but not high enough to lead to more WH.

Your mention of WH does invalidate somewhat my very very napkinny math above, as WH means the initial 120 focus don't translate directly into 4 basic attacks, but more like 2 WH attacks and then a pause.

NoGoal · 09/11/12, 12:13 PM

Originally Posted by Nooska

Completely basic, to not delay pet basic attacks, you need a regen of 30 focus per 3 seconds

Just noticed that too. Pet basic attack is 25 focus so it would ´only´ require 59% crit chance with 0% haste and 32% with 20%. Or am I missing something?

[side note: In theory, it's possible to cap pet focus during Blood Lust+Rapid Fire with 30% haste and 38% crit chance. I doubt it's worth (or even possible) losing so much mastery for haste.]

Nooska · 09/11/12, 2:34 PM

Indeed the basic attack only costs 25 focus, so we need 25 focus per 3 seconds, or 8.333 (repeating) fps - lets round that to 8.4 for sanitys sake. The pet stil contributes 5 fps at 0 haste, leaving 3.4 fps from GftT per second. This becomes 10.2 focus per 3 seconds - or actually, a very clean 10 focus per 3 seconds (15 from pet, 10 from GftT), leading to a required 0.666 (repeating) GftT procs per 3 seconds. At 1 autoshot per 3 seconds that means a sustained focus regen requires a 66.7% crit rate.
Looking at the previous 5 minute fight, and ignoring WH for the moment, we are looking at slightly less of course, due to the starting 120 focus; skipping the math, we indeed need the 59% (or accurately 58.7%) crit for GftT to supply enough pet focus for 1 basic attack every 3 seconds.

At the 20% haste (indeed and arbitrary umber) we would require 38.9% crit for sustained pet focus regen, or the 32% you correctly state.
If FF increased our and our pet focus regen (I don't believe it does, but will go check in a minute), we would have pet focus sustainability under FF with "merely" 27 % crit and raid buffs.
During RF and and BL we will have sustainable pet focus regen with respectively 19% crit and 28.2% crit respectively - not factoring in raid buffs, so in practice 14% respectively 24% crit.

This does influence when we should use RF in the opener sequence, namely at least 3 seconds after our pet is on the target, making it our 4th (or later) GCD, with the starter being BW, GT, KC(BS),LR/MoC(KC),(DB), RF, Readiness.
With the value RF gives to pet focus regen combined with GftT, it should definately be used before readiness. The question is whether it will be most advantageous to save the next RF till the third BW (1 minute after the second BW). My gut feeling is its best to stack with BW now (or just prior to, to get up pet focus) for sustaining pet focus during BW, and the 3 minute CD will be up before Readiness in any fight where we get 2x Readiness, so we will get the same 2xRF in between Readiness as we would if we popped it immediately.

Lokrick · 09/11/12, 10:23 PM

Originally Posted by Nooska

...This does influence when we should use RF in the opener sequence, namely at least 3 seconds after our pet is on the target, making it our 4th (or later) GCD, with the starter being BW, GT, KC(BS),LR/MoC(KC),(DB), RF, Readiness.
With the value RF gives to pet focus regen combined with GftT, it should definately be used before readiness. The question is whether it will be most advantageous to save the next RF till the third BW (1 minute after the second BW). My gut feeling is its best to stack with BW now (or just prior to, to get up pet focus) for sustaining pet focus during BW, and the 3 minute CD will be up before Readiness in any fight where we get 2x Readiness, so we will get the same 2xRF in between Readiness as we would if we popped it immediately.

Simc and FD both showed it to be a win to be under rapid fire for stampede. The increased focus then helps all the pets to get in another basic attack before vanishing. Simc was showing it to be worth delaying the second round of rapid fire until the second set of stampedes (at 5 minutes and change) if the fight was going to be long enough. There have been just enough changes since that initial analysis (e.g., rabid nerf) that they should be re-run, but that delay is worth considering.

Nooska · 09/12/12, 3:50 AM

With readiness and stampede both being 5 minute cooldowns, I would think that RF would be up automatically with stampede, since we could simply delay stampede for the few seconds it t akes to get the prereadiness stuff going, then readiness and stampede+RF.
I am curious as to how SimC gets RF to be a benefit for stampede though - RF is ranged haste, the stampede pets don't use focus (do they? I haven't seen any indications of it in beta) - so what affects stampede in RF.
(All that said, I clearly forgot Stampede up there in my opening)

Rivkah · 09/12/12, 4:05 AM

Originally Posted by Nooska

With readiness and stampede both being 5 minute cooldowns, I would think that RF would be up automatically with stampede, since we could simply delay stampede for the few seconds it t akes to get the prereadiness stuff going, then readiness and stampede+RF.
I am curious as to how SimC gets RF to be a benefit for stampede though - RF is ranged haste, the stampede pets don't use focus (do they? I haven't seen any indications of it in beta) - so what affects stampede in RF.
(All that said, I clearly forgot Stampede up there in my opening)

Stampede pets work just like our normal pets. They start out at full focus, have all the same talents including wild hunt and the same basic regen mechanics including inheriting the hunter regen multiplier (they don't get the attack speed from RF but they definitely get the regen). They don't benefit from things that are designed to affect our active pet, like focus fire, cobra strikes, go for the throat and fervor (although they do stack frenzy for themselves, you just can't consume it with focus fire).

It occurs to me that although I currently don't have stampede pets modeled to proc invigoration, they probably can (I haven't tested it, but if they can stack frenzy it stands to reason they can proc invigoration).

So yes, RF is a big benefit for stampede given that it can make the difference not only in the total number of basic attacks they do but also whether they count as wild hunt. I also haven't modeled it, but it's possible that forcing focus fire to activate regardless of stack count before (or right after) casting stampede may be beneficial as well, so that your stampede pets will have more focus regen during the short time they're active.

Exhawk · 09/15/12, 10:43 PM

Breakdown: Lynx Rush vs Blink Strike

First let's look at the spellbook:

Blink Strike:

Instant	20 sec cooldown
Causes your pet to instantly teleport behind an 
enemy target up to 40 yards away from your 
pet and inflict 600% normal damage.

Lynx Rush:

Instant	1.5 min cooldown
Your pet rapidly charges from target to target, 
attacking 9 times over 4 sec, dealing 200% of 
its normal attack damage to each target.

And the 2 main abilities that will be linked/tested with these:

Bestial Wrath: 20% Damage boost - 10 Seconds - 1 Min CD
Rabid: 25% Pet Damage - 20 Seconds - 1:26 Min CD

So let's test the damage in a 3 min period, since everything comes off CD and lines up after 3 mins as far as these spells go.

In a 3 min period, without casting any of them as they come off CD at 3 mins, you get:

9 Blink Strikes
2 Lynx Rushes - 9 Attacks each
3 Bestial Wraths
2 Rabids

Spreadsheet with Breakdown/CD's: http://i.imgur.com/hcc9x.jpg
Spreadsheet with Line-up: http://i.imgur.com/SbdsS.jpg

The math for BS goes:
(600 x 5 unbuffed damage) + (600 x 1.45 x 2) + (600 x 2 x 1.2) + (600 x 1.25) = 6930% total damage

The reason for (600 x 1.45 x 2) is 600% damage multiplied by the 45% buff from BW and Rabid, multiplied by 2 for readiness usage at the start of the fight, and only if used while during Rabid (not shown in chart)

Without Readiness at start, you lose one of the 600 x 1.45, so you're left with 6060%

The math for LR goes:
(200 x 9 x 1.25) + (200 x 9 x 1.45 x 2) = 7470% total damage

The reason for (200 x 9 x 1.45 x 2) is 200% damage multiplied by 9 attacks over 4 sec, multiplied by the 45% buff from BW and Rabid, multiplied by 2 for readiness usage at the start of the fight, and only if used while during Rabid (not shown in chart)

Without Readiness at start, you lose one of the 200 x 9 x 1.45, so you're left with 4860%

With Readiness at start after initial BW, BF, Rabid, DB, and BS/LR.

LR (7470%) > BS (6930%)

Lynx Rush pulls ahead by a small margin (7.7% difference).

Without Readiness

Blink Strike (6060%) > Lynx Rush (4860%)

Blink Strike pulls ahead by a wide margin. (24.7% difference).

Lilbitters · 09/15/12, 11:44 PM

Originally Posted by Exhawk

Breakdown: Lynx Rush vs Blink Strike

The case isn't quite this simple as you also have to consider the GCDs use for Lynx Rush and Blink Strike during the 3 minute time frame and the benefit gained by the Lynx Rush scenario of being able to cast more Arcane/Cobra Shots.

Using a FemaleDwarf T14N profile set to a 3 minute duration and using the following minimal shot selection:

Bestial Wrath
A Murder of Crows / Lynx Rush / Blink Strike
Readiness
Arcane Shot
Cobra Shot

69578.72 = A Murder of Crows
66857.77 = Lynx Rush
66755.60 = Blink Strike

Although it is true that a larger percentage of damage will be done from Blink Strike than Lynx Rush, the net damage will be large by having more GCDs available for other abilities when using either of the alternatives.

Exhawk · 09/16/12, 12:39 AM

Originally Posted by Lilbitters

The case isn't quite this simple as you also have to consider the GCDs use for Lynx Rush and Blink Strike during the 3 minute time frame and the benefit gained by the Lynx Rush scenario of being able to cast more Arcane/Cobra Shots.

Using a FemaleDwarf T14N profile set to a 3 minute duration and using the following minimal shot selection:

Bestial Wrath
A Murder of Crows / Lynx Rush / Blink Strike
Readiness
Arcane Shot
Cobra Shot

69578.72 = A Murder of Crows
66857.77 = Lynx Rush
66755.60 = Blink Strike

Although it is true that a larger percentage of damage will be done from Blink Strike than Lynx Rush, the net damage will be large by having more GCDs available for other abilities when using either of the alternatives.

So the difference is smaller than I thought. If you pair that with how Blood Fury and Rabid line up, LR takes a bit of a bigger lead, which makes my math semi correct.

The difference on FD sim without Rabid/BF line up is 0.15% difference in total overall damage, Mine was 0.5% with Rabid/BF.

Clearly Murder of Crows is the leader though as far as DPS on single target goes.

I just wanted to settle a debate many of the people on the regular WoW hunter forums were having.

Glatzel · 09/21/12, 4:17 AM

I don't think that those numbers are completely trustworthy, because lynx rush and blink strike both profit from rabid (wich should be in that simulation, because every lynx rush benefits from it and blink strike does, too, once in a while) and i don't know if AMoC does and in a fight with exactly 3 minutes length you only get 3 lyns rush, even with readiness. would not be 3 minutes and 8 secounds more accurate because the cd of lynx rush is at 4 seconds when the simulation ends.

edit: from my testing, AMoC does not benefit from rabid, so it should definitely be in the simulation.

Lilbitters · 09/21/12, 8:07 AM

Originally Posted by Glatzel

I don't think that those numbers are completely trustworthy, because lynx rush and blink strike both profit from rabid (wich should be in that simulation, because every lynx rush benefits from it and blink strike does, too, once in a while) and i don't know if AMoC does and in a fight with exactly 3 minutes length you only get 3 lyns rush, even with readiness. would not be 3 minutes and 8 secounds more accurate because the cd of lynx rush is at 4 seconds when the simulation ends.

edit: from my testing, AMoC does not benefit from rabid, so it should definitely be in the simulation.

I did retest this as I had some concerns about the numbers myself and wanted to recheck. Someone can feel free to test these in both SimulationCraft and Esoth's Hunter Spreadsheet for more verification.

Starting with the current BM Max DPS profile in FemaleDwarf and modifying it with the following:
Removed from Shot Priority: SrS, KS, KC, RF, DB, GT, Stamp, ET, AS
Changed Fight Length (in minutes): 7.5 -> X

If used on cooldown, the expected cast times for each ability within 7.5m should be:
A Murder of Crows -> 0s, (Readiness) 30s, 150s, 270s, (Readiness) 300s, 420s
Blink Strike -> 0s, (Readiness) 1s, 21s, 41s, 61s, [...], 281s, 301s, (Readiness) 302s, 322s, [...], 422s, 442s
Lynx Rush -> 0s, (Readiness) 4s, 94s, 184s, 274s, (Readiness) 300s, 394s

The best time to compare the abilities appears to be just before the second Readiness at 5 minutes (300s), so I ran the modified profile at each time setting and compared the Combined DPS to the highest ability at that time setting.
*Note: The case for Blink Strike is exaggerated as the only ability it is replacing is Cobra Shot, when it reality it will lose more DPS because of the higher opportunity cost associated with delaying itself or other higher DPS abilities.

Combined DPS At 5.0 minutes
065405.06 (+00000.00) ~ A Murder of Crows
063415.54 (-01989.52) ~ Blink Strike
063364.94 (-02040.12) ~ Lynx Rush

Below is a more detailed breakdown at other various time settings:

← Click Here

Rivkah · 09/21/12, 11:39 AM

It's not surprising that the relative value of Lynx Rush to Blink Strike changes based on the duration of the fight, relative to how close to a Readiness cutpoint you get. Readiness is a big factor in why Lynx Rush is more attractive than Blink Strike- because a higher damage ability gives you more bang for the readiness buck.

Keep in mind that my site will average the benefit of Rabid over the course of the fight, so it also doesn't completely reflect the extra value you get from Rabid's cooldown lining up almost perfectly with Lynx Rush for BM (or Beast Within being up for alternating Lynx Rush casts). Simulationcraft models this more accurately and so you see Lynx Rush ranking higher in there.

In actual play, using Blink Strike perfectly will be more difficult than Lynx Rush since it has to be used more often. However Blink Strike does provide additional benefits which will suit some fights well, whereas Lynx Rush may be more attractive when AoE targets are involved. Lynx Rush also can lead to parries on other pet abilities so there are risks involved in the timing of it (i.e. using it when you're about to kill command might not be ideal). Murder will be the easiest to get consistent results from as it has the longest cooldown and doesn't have to be lined up with other abilities to get full value. So which talent best for players who execute perfectly may not match which is best for players who might be a bit less meticulous.