Well, according to WoWWiki pets will tick for 25 focus every 4 seconds. [EDIT]Glau has corrected this, it's actually 24.5 focus per 4. The remainder of this post will be updated to reflect that. So that's 6.125 focus per second and each point in Bestial Discipline adds 3.0625 focus per second. GftT adds
Focus from GftT / second = shots / second * crits / shot * 25 * talent points in GftT
Note that crits / shot is the same as the crit rate. Using my Sunfury bow, I'm at 2.10 attack speed, so with steady shot I'm doing shots every 2.1/2 = 1.05 seconds (note that I'm ignoring iAotH), so I'm doing 0.95 shots / second. Assuming the Sunfury is 'the' bow to use currently (or rather, the bow that the majority of raiders are using) for the purposes of speed, we can optimize the talent distribution between the two talents based on crit rate. Rather than looking at it as "talents in one versus talents in the other" I'm going to assume that we have a pool of 2 talent points (the other two being moved to imp mend pet) and any points NOT in GftT would be in Bestial Discipline. It's a simple optimization problem now.
let
x = talent points in Bestial Discipline
y = talent points in Go for the Throat
x + y = 2
F = focus
dF/dt = 6.125 + 3.0625 * (x) + 0.95 * Crit Rate * y * 25
Set dF/dt = 0 and solve for x
0 = 6.125 + 3.0625 * (x) + 0.95 * Crit Rate * (2 - x) * 25
-6.125 - 2 * 23.8 * Crit Rate = (3.0625 - 23.8 * Crit Rate) * x
x = (-6.125 - 2 * 23.8 * Crit Rate) / (3.0625 - 23.8 * Crit Rate)
So for a 20% crit rate the maxima for focus regeneration occurs when you have 6.8 points in BD. Clearly this isn't realistic as it's outside the boundaries (that is, x + y = 2 and 0<= 2 - x <=2 and 0<= 2 - y <=2). Lets just plug in integer values for x and y and see what the focus regen looks like.
Perhaps the result was immediately apparent to everyone else, but I'm a little surprised that GftT is superior by such a wide margin on a point for point basis relative to Bestial Discipline for any reasonable amount of crit that a raiding hunter would have.
As for pet focus usage and how to spec for it, if you're using a ravager or cat autocasting gore/claw and bite then you'll be using 16.67 fps for gore/claw and 3.50 fps for bite, for a total of 20.17 fps. Even with a 25% crit rate, you can't keep up with the focus usage of your pet. If you're at 25% to crit, 1 point in BD will keep your pet fat and happy, in terms of focus. If you're at 20% to crit, 2 points are required to keep your pet contstantly autocasting.
Lets find the equivalence points for needing points in BD. If you have 2 points in GftT and 1 point in BD then you'll just be keeping up with your pet's focus usage (again, assuming ravager/cat) at 23.1% to crit. If you have 23.1% to crit or more, 1 point in BD keeps up with your pets focus usage; less than 23.1% to crit and you need a second point in BD to keep up with the focus usage. Likewise, if you have 29.5% or more to crit, you don't need any points in BD to keep up with a ravager's/cat's focus usage.
If you're using a windserpent spamming LB (no bite) you're using 33.3 fps (50/1.5 GCD). To keep it up with a constant spam, you'd need a 57.3% crit rate with no points in BD. With 1 point in BD you'd need a 50.8% crit rate and with 2 points in BD you'd need a 44.3% crit rate. It's very difficult to keep up with a wind serpent casting LB every time the CD is up, clearly.
Just some food for thought regarding pet focus usage.
Last edited by The Iron Colonel : 05/11/07 at 4:10 PM.
Pet focus regen is actually about 24.5 per 5 (or there is a small amount of focus decay). It is very easy to tell if you use a wind serpent. After the first two LBs your pet will have 48-49 focus after the first regen tic and can't cast another LB till the next tic takes it up to 96-99.
Edit regarding GftT:
It has always been a very powerful talent and is the sole reason LB got nerfed. Without GftT the scaling LB wasn't that powerful.
Edit 2:
Modeling GftT as a per tic amount of focus regen really overvalues it. You lose a fair amount to back-to-back crits, crits happening during regen tics, etc.
Pet focus regen is actually about 24.5 per 5 (or there is a small amount of focus decay). It is very easy to tell if you use a wind serpent. After the first two LBs your pet will have 48-49 focus after the first regen tic and can't cast another LB till the next tic takes it up to 96-99.
If that's the case, it just pushes the balance more in favor of GftT over BD, and I'll correct my post to reflect a 24.5 focus / 5 innate regen rate. I was mostly engaging in a little math to figure out which configuration of BD/GftT would keep a BM pet stocked on focus if you didn't have both maxed (as in the case of Howitzer's last post, where he posits removing two points from BD and placing them in Imp Mend Pet).
[EDIT]
Originally Posted by Glaurong
Edit regarding GftT:
It has always been a very powerful talent and is the sole reason LB got nerfed. Without GftT the scaling LB wasn't that powerful.
You're probably correct there - I recall seeing theoretical numbers putting LB spam over 100 dps as it previously scaled.
Last edited by The Iron Colonel : 05/11/07 at 2:13 PM.
-No fel str, just onslaugh
-No food buff
-No flask
-No warrior in group
Shame he's immune to FF.
Live servers btw.
What I would like to know, how much is the armor reduction on Hydross? Based on your gear, your rotation, group set up and pot usage I'd say he cant have more then 15%. Is that about correct?
Edit: btw, whats also sick, is the dps of your shammy...
Welshy, you may have been correct about pet regen ticking every 4, rather than 5. I'm not positive on that part. The main thing I was addressing is that it wasn't 25 per tic.
Whether the tics happen every 4 or 5 I'm really not sure.
Nice going Osse. I must say BM is looking hard to pass up. As a hunter class lead and guild officer I dont think I am going to be able to however since I want at least one EW hunter since our raid makeups tend to be phys dps heavy (cant seem to recruit reliable mages that pump the dps) and i don't see that being anyone else but me.
I am hoping at least one of our hunters does try it and will enjoy seeing him top the charts and live vicariously through him
------------------------------------------------------------------
Not quite sure where to put this, since the old thread is 90 pages long and not quite pertinent for 2.1 anymore. I did some messing around on the live server in Blasted Lands with my new pet Scorpid. I found some pretty intriguing stuff. From my testing, here is what I think the formula is for Scorpid Poison (all testing based off of Rank 5, which is different in duration from all other ranks of the spell):
Initial Stack Damage Per Tick= ( ( (0.125*rAP) + 44)/ 4) * % Damage Modifiers
Damage Per Additional Stack = 11 x % Damage Modifiers
Damage Per Tick = ISD + ((# of Stacks-1) * DPAS)
TOTAL Damage of Scorpid Poison (Rank 5) = (ISD + DPT) * 4
Notes:
-If this seems confusing, just think about it this way: The first time your pet casts Scorpid Poison, every tick of the poison will do Initial Stack Damage Per Tick. Each additional application of Scorpid Poison (it stacks 5 times) will increase the amount of damage done by the Damage Per Additional Stack. So if your ISD is 50, and your DPAS is 10, then at the third stack each tick will be doing 70 damage (50 + 10 + 10).
- Every successful application of Scorpid Poison will maintain the same debuff the entire time, stacking up to 5 times. Essentially, if your pet fails to apply Scorpid Poison twice in a row, that debuff will disappear and you will have to start all over again from scratch with the ISD. As I will explain later, this isn't always a bad thing.
- "% Damage Modifiers" are the additive total of any buffs you might receive that increase your pet's damage. This includes the BM talents Unleashed Fury, Ferocious Inspiration, and Bestial Wrath; I would assume this also includes non-self buffs as well. So for example, if you were being buffed by all three aforementioned BM talents, your multiplier would be 1.48 (3% FI + 25% Bestial Wrath + 20% Unleashed Fury). However, it appears that ISDPT gets no benefit from the passive 20% modifier you receive from Unleashed Fury. I'm not quite sure why this is the case. Either it isn't receiving a buff it should, or there is something off with my math (rAP->spelldamage conversion, % of spelldamage translated into DoTs, what have you).
- Damage Per Additional Stack is only modified by % Damage Modifiers. This is important; if you only have AP buffs at the time, the only increase to Scorpid Poison damage you will see is in the ISD. You are potentially missing out on what can be as much as 30 DPS if your pet is able to maintain the same stack of Scorpid Poison throughout.
- ***EVERYTHING IS ONLY CALCULATED ONCE, UPON THE FIRST APPLICATION OF SCORPID POISON.***
This is very, very important. If you hit Bestial Wrath when your pet has a full 5-stack of Scorpid Poison already up and running, it will not gain any benefit from the talent. Hell, even if it only had 1 stack before you hit the button, it will not gain any benefit from your new buff. As long as that same stack is being kept up by your pet, every tick will do the same damage it did before. Conversely, if you hit all your cooldowns BEFORE any Scorpid Poison has been applied, then your pet casts the spell, as long as that debuff is kept up your pet will get it to the maximum damage that was calculated in the beginning, even if all those buffs you had drop the second after the first debuff was applied.
Conclusions:
Based on the above, what you want to do is get every single AP buff and Damage Modifier buff you can get all at once, wait a few seconds for the rAP to get translated into your pet's spelldamage, THEN send your pet in (or turn on Scorpid Poison). I managed to buff my pet's DPT from 132 to 230 just by doing the above with my own self-buffs and cooldowns. I'd imagine under a raid setting this number can jump even higher.
Scorpid Poison's initial cost is actually relatively cheap compared to other pet skills. If all you did was get one stack of Scorpid Poison throughout, you would get at least 8 Damage Per Focus, which is a huge increase over the other talents. More specifically: if all it did was tick once a la a "direct damage" ability, you would get 2 DPF, which is on par with Gore and Claw as a pure Focus dump (the only thing stopping it would be the 4 second cooldown). If you looked at Scorpid Poison on a per-tick basis, this number jumps up astronomically when you include a full 5-stack debuff that is being maintained. Consider a Scorpid Poison debuff that is ticking for 250 damage every two seconds - this translates into 125 DPS (to compare, full-bore LB spam that hits for 200 a pop is 133 DPS). On a good day, Gore will give you an extra 30 DPS, assuming no misses/glances/blocks/parries/dodges/armor mitigation. This is damage that can only be matched by Lightning Breath when it has no Focus concerns (i.e. spamming LB every cooldown), and comes at a Focus cost that is more than 4.4 times cheaper per single tick, and almost 18 times cheaper when allowed to run its full course and never reapplied. If your pet is able to maintain a full stack debuff for any length of time, the cost of this ability scales geometrically low in terms of FPD (15 + 7.5 * (# of Stacks - 1)).
Of course, these are all theoretical maximums. What will probably end up happening, even with the changes to melee combat mechanics, is that your pet will miss some applications, and will allow the fully stacked debuff to drop a handful of times in any raid boss fight. However, when you consider that there will be times where a fully buffed, fully stacked debuff can be maintained for a reasonable period of time (or even just a regular old self-buffed fullstack), I'd bet good money that Scorpid Poison would handily beat any Gore or pure-Claw pet, and could probably match or even beat Lightning Breath on a good day. Yes, even with the 4-second cooldown.
The thing I'm not mentioning here is that unlike all the other spelldamage-based pet abilities (all of which except LB are "awesome" when you look at them in pure math terms), SCORPIDS GET CLAW TOO. No other pet has a combination of Focus Dump + Scaling Ability. Some pets get Claw/Bite or Gore/Bite or even Gore/Charge, none of which scale. Or Bite and Poison Spit/Fire Breath, both of which are on long cooldown timers. So while your scorpid pet is applying Scorpid Poison every 4 seconds at a measly 30 Focus cost, it's using all that excess Focus on Claw, which is still a pretty nice Focus Dump (Gore is only slightly better - what makes Ravagers so good is their innate DPS bonus combined with Gore, not just the ability itself). The two skills combined use a little over 24 FPS, which as previously shown in this thread is an easily sustainable number.
Long story short - I think Scorpids rule, and if I were still raiding I'd try mine out on some raid bosses. The innate DPS hit you take from using a Scorpid would be easily countered by the fact that you can sustain a >100 DPS ability on a consistent basis (consider if your pet does 300 DPS total, a 10% modifier is only 30 DPS, and a 20% modifier difference - the difference between a Ravager and a Scorpid - is 60 DPS). The reason I'm sharing this with you guys so "soon" is because I really don't have the opportunity to do real-world raid-testing anymore, so I'm hoping someone will take the initiative and tell me if the theorycrafting holds up.
*It appears that the increased rAP from Aspect of the Hawk is not translating into spelldamage for the pet. If someone can confirm this with a Windserpent or another pet that would be awesome. It took me about 15 minutes to figure this out, because it was skewing some of my numbers. Once I took out the Aspect of the Hawk buff, everything fell into place.
If the 17% resist rate for a boss mob applies to your pet the same way it applies to you then I would imagine that throws a major kink in this. What do the numbers come out like if you assume a 17% resist rate for each poison application?
Edit: I also believe you are calculating the added spell damage incorrectly.
Unless they apply your RAP to spell damage differently for different classes of pets a scorpid would only get 3% of your AP as spell damage (this is what testing of LB shows).
Since scorpid poison is a dot, each tic gets 1/5th of your overall spell damage (might be different here because it is ticking every 2 seconds instead of 3 like most dots). Which means 0.6% (0.006) of your total AP gets directly added per tic, per stack. So a full 5/5 stack will get 3% of your AP applied directly as damage a tic.
If the above is right then with 2000 RAP, a 5/5 stack should tick for 115 every 2 seconds before modifiers. With pet family and happy modifiers 115 * 0.93 * 1.25 = 134 every 2 seconds
If somehow the scorpid is still getting 12.5% of your AP as spell damage then 12.5% of your AP would apply directly as damage a tic.
So in this case with 2000 RAP a 5/5 stack will tick for 305 every 2 seconds before modifiers. With pet family and happy modifiers 355 every 2 seconds.
Edit 2: Never mind, I see you based your formula off tests. I guess I didn't read close enough.
Wow, that's some good news. Most bosses are not immune to poisons (notable exceptions like Hydross not withstanding), so that's cool to hear. Likewise, if you're raiding with mutilate rogues they'll never have to worry about shiving to put up poisons, enabling them to spend their energy on other abilities.
Couple questions: do you have any information on the miss rate / resist rate of poison? If you do, it would be fairly trivial to determine the average up-time on poisons, which would be a nice data point to have. Also, are you seeing any partial resists on poison ticks right now? Again, useful information against which to compare other pets.
One of my primary concerns (albeit totally unsubstantiated and speculative right now) would be prioritizing claw versus poison. Currently I'm using a cat with claw/bite; obviously I see many more claws than bites, but what would happen if you hit a drought of crits and your pet doesn't get a poison stack renewed? Seems like you might lose all the bonuses from cooldowns/trinkets that you've spent on the stack.
I'll pick up a scorpid and try to level it a little bit over the next few days; I don't have any raids scheduled until Tuesday, and we're still farming Karazhan/Gruul for gear, so my data my not be representative of leading edge raid encounters.
@ Glaurong:
You're right, I mistyped the 5 in the beginning. All of the math was done for regen ticks every 4 seconds, but I had 5 listed in the documentation; I corrected it. Thanks for the heads up!
[EDIT] Glau beat me to it with the resist figure. If you assume that you'll get 2 chances to renew the stack (8 second duration, 4 second CD on poison) you're looking at 1-0.17^2=0.9711 chance to renew the stack; if you assume 1 chance to renew, it's a 1-0.17=0.83 chance to renew. Can you have a poison miss/dodge/parry result? If so, that will also reduce the uptime on poison. Is the poison stack shared by more than 1 scorpid? That would increase uptime substantially, but return less overall damage as the stack is limited to 5 applications. I'll check to see if more than 1 scorpid can share a stack this afternoon.
Last edited by The Iron Colonel : 05/11/07 at 4:35 PM.
If the 17% resist rate for a boss mob applies to your pet the same way it applies to you then I would imagine that throws a major kink in this. What do the numbers come out like if you assume a 17% resist rate for each poison application?
They won't look as good on a pure DPS standpoint, but no one ever looks at the miss rates for Gore/Claw and the resist rate for Lightning Breath, either. I agree though, it is something that needs to be considered. But like I said, even in a worst-case scenario and your pet only hit every third application and was only able to sustain a single-stack debuff of Scorpid Poison, the damage would be pretty close to current focus dumps because you're getting a DoT with base damage that almost singlehandedly matches other focus dumps on a pure DPS basis anyway (yes, even with just one stack). The fact that it would be spending the rest of its time spamming Claw pretty much guarantees that the end result is better than a pet spamming just Claw or Gore. You're not missing out on much in terms of opportunity cost when you're replacing Claw with Scorpid Poison 37.5% of the time, especially when one of them is a DoT (meaning the damage from one doesn't completely replace the damage from the other). Most of the time you are hitting Claw at no loss in opportunity cost whatsoever because SP is on cooldown and/or already applied, and you are still getting the full damage from that ability while hitting Claw.
Unfortunately, with stacking debuffs like this, it is impossible to properly model the amount of damage done because the accurate model is a series of curves rather than just one straight number or series of numbers.
[Edit Add]
I based my formulae on my own personal testing in Blasted Lands. I'm 99% certain my numbers are right. With ~1600 rAP my pet was seeing initial applications of 62 damage per tick. This translates into 248 damage over the life of one stack, for a 204 damage increase over the initial 44 damage listed on the tooltip. With another 270 rAP from Bloodlust Brooch, my initial stack was hitting for 69-70. The real-world math adds up. Are you sure you're reading my formula properly? It's damage per tick, not overall damage. The number eventually gets divided by 4 at the end. 1600 * .125 = 200. 200 + 44 = 244. 244/4 = 61. Or am I missing something?
Keep in mind they only nerfed Lightning Breath's spelldamage modifier. They probably ignored the other spells beacuse they're all on relatively long cooldowns, so your pet can't spam them anyway.
*Actually, you may be partially right. I'll have to do some double-checking, but either pets don't get a benefit from Unleashed Fury to their initial stack or I'm missing something in the conversions. Either way, LB is a bad spell to base rAP->spelldamage conversion off of, because Blizzard treated it as a special case and nerfed it to hell.
TIC:
There is a level 69 Scorpid in Shadowmoon Valley that has Scorpid Poison (Rank 5). It took me a few hours to tame him, and get him to 70 with full Loyalty (the former will happen before the latter).
I based my formulae on my own personal testing in Blasted Lands. I'm 99% certain my numbers are right. With ~1600 rAP my pet was seeing initial applications of 62 damage per tick. This translates into 248 damage over the life of one stack, for a 204 damage increase over the initial 44 damage listed on the tooltip. With another 270 rAP from Bloodlust Brooch, my initial stack was hitting for 69-70. The real-world math adds up. Are you sure you're reading my formula properly? It's damage per tick, not overall damage. The number eventually gets divided by 4 at the end. 1600 * .125 = 200. 200 + 44 = 244. 244/4 = 61. Or am I missing something?
Keep in mind they only nerfed Lightning Breath's spelldamage modifier. They probably ignored the other spells beacuse they're all on relatively long cooldowns, so your pet can't spam them anyway.
Yeah I misread you the first time through, didn't pick up that it was based on testing. Scorpid maintaining the 12.5% makes my testing with LB make a lot more sense, they just gave LB a 0.1 coefficient.
Yeah, I'm doing some testing with Pet Happiness (GG completely forgetting about that one), and I'm not quite sure how it fits in. At this point, I'm not quite sure if the formula is correct overall. But assuming your pet is Happy and you are BM-specced, I'm comfortable saying that the formula will be accurate for you. There's something weird going on with the scaling damage modifiers, but other than that everything seems ok. If I try to adjust the rAP->spelldamage or the spelldamage->IDPT numbers, they get very skewed off when compared to the real-world numbers I have from my own testing. However, when Happiness is adjusted, the numbers change ~25% from what they were, so it seems that there is some strange math going on in the middle that isn't being calculated in any way that I can figure out. I'm guessing there is a combination of damage multipliers that are being added together, with some like Happiness being left out until the very end of the calculation.
Update: Scorpid poison can be dodged/parried/miss, which will have a large effect on it's uptime. I've been testing it on boss mobs in AV (which should be level ~73).
Sorry for the delay on this reply. Thanks for the feeback, I'll keep improving the dps when I get back to Finland. Currently typing this from Budapest and will be here for a few days.
About my mana pool and mana use, yeah it's expensive. :>
Very doable to keep up the rotation for a long period of time with fel mana pots and a few judgements.
About the rotation, I've tested many bow speeds and 3.0 turned out to work pretty nicely, although on this Karazhan clear Phoenix bow finally dropped and got 15 dps increase when I tested dps for 15 mins in Blasted Lands.
I'm not sure why almost pure auto/steady rotation works for me, I always have close to 50 ms during boss fights and 60fps. the gear I'm using right now is the gear I built for survival testing which also turned out to work well with Wolfslayer but dps just was higher with BM so decided to stay as BM for a while.
I had a scorpion for a while when I was MM and could keep it stacked 5 times on most boss fights but if there's dps breaks etc it just doesnt work that well in my opinion. And what bothers me as well is that it doesnt trigger feral inspiration. I like this ravager (which I leveled in two days) :p, although one problem is that it cant burn focus with 37% crit sometimes but there's not really any other options to fill those two points in BM tree and you'll always get unlucky rows where you just dont crit, hence why I havent dropped the talent yet.
What I would like to know, how much is the armor reduction on Hydross? Based on your gear, your rotation, group set up and pot usage I'd say he cant have more then 15%. Is that about correct?
Edit: btw, whats also sick, is the dps of your shammy...
Not sure about his armor but I've reached 1250 on Mag when clicking cubes. :p
Mag DPS can be a bit misleading because of the cube clicking. When you have to do it, you aren't DPSing. When you don't, your DPS is dependent upon how long the other people are willing to channel for the extra damage debuff.
Mag DPS can be a bit misleading because of the cube clicking. When you have to do it, you aren't DPSing. When you don't, your DPS is dependent upon how long the other people are willing to channel for the extra damage debuff.
Sorry for asking, but I havent been to Mag yet. So this cube clicking is some damage enhancing gimmick?
I'm always more of a fan for dps comparisons on straight dps fights with almost no movement, no dps enhancements and for some minutes at least.
So as simple an enemy as he might be, I find that Attumen is a good base for comparisons, because he has regular armor, no gimmicks, everyone does it and its basically tank'n spank.
Reaching over 1000 dps on him would really get me looking close then.
My best was around 980 dps in a WWS log, but I had my perfect group with feral, enhancement and warrior.
Sorry for asking, but I havent been to Mag yet. So this cube clicking is some damage enhancing gimmick?
I'm always more of a fan for dps comparisons on straight dps fights with almost no movement, no dps enhancements and for some minutes at least.
When he is being banished, he takes a large % more damage (not sure on exact number), however, at least in my guild, we only spend the bare minimum time clicking the cubes, in order to reduce the healing needed on the raid. Sure this means 1 or 2 minutes extra on the fight, but we manage.
Originally Posted by Ulthwithian
Paladins do have an ability to heal multiple people at once. It's called Divine Storm. ><
*Actually, you may be partially right. I'll have to do some double-checking, but either pets don't get a benefit from Unleashed Fury to their initial stack or I'm missing something in the conversions. Either way, LB is a bad spell to base rAP->spelldamage conversion off of, because Blizzard treated it as a special case and nerfed it to hell.
Gore,
is it possible that 'what's missing' is the -6% dmg modifier all Scorpids have?
I didn't see it included anywhere in your math in the first post.
05/11/07, 11:32 AM
Similar Threads
