Thanks for the spreadsheet link. Using that, and using reasonable mean values for auto, steady, and kill command damage (i.e., mine from Cheeky's), and using a fixed dps weapon and changing speed only, it turns out that the highest total damage per 1000 seconds comes from a 2.8 speed weapon with the 3:2 macro using kill command and spammed 20 times a second.

Yes, lots of caveats. But it is interesting to see the damage peak at 2.8 (though 3.0 is close.)

Note I am using particular weights for auto, steady, and kill command. Your mileage will vary. But as long as the values were reasonable (I tried both a custom weapon as well as WSR), the peak still occurred at the same place.

Here's my table and the equation I used:

This table appears to show a higher DPS for 1:1 than 3:2 with a 2.7 speed weapon, however we know this to not be true. Am I misinterpreting something?

The question wasn't directly answered. With 20% less DPS the DPS of a boar doesn't come close to a Ravager, Cat or Wind serpent.

A wind serpent has 2% more health than a cat, and no abilities that would affect the survivability. If you aren't keeping a cat up right now, bringing a wind serpent isn't going to make any difference.

Cool, thanks very much. Guess I'll keep bringing my ravager to raids but will keep my Agamar for stuff like heroics and farming. He's just far too cute, and I've worked too hard to level him from 25 just to dump him because he's doing less dps.

Hey guys... I just had a couple of quick questions and maybe you'll have the answers or I can at least be pointed in the right direction to find them.

1. When does ignore armor begin to shine for hunters PvE wise? Is there a set amount to shoot for to where it'll out perform straight AP?

2. Will stacking ignore armor over AP hurt our pets more than helping us since it seems our pets scale with our AP and the ignore armor stat doesn't benefit them.

3. How can you find out what your pet's crit rating is? Does it scale with gear or is it a base amount that can be added to with talents?

No misinterpretation on your part. The numbers are what they are. It's interesting he had ~50msec higher lag during the 1:1 testing compared to the 3:2 yet got better dps. If only hunter shot rotations were easy to analyze...

Archimonde is a bad fight for pets. Even with max FR, doomfire's dot just tears them up; a mend just isn't enough to heal the damage. With a shadow priest, 2 pc T5, or both, you might be able to out-heal it. You may want to consider just keeping your pet with you so that you have the 2% damage from Focused Fire, and have TBW for a fear prevention. As you get used to the fight you'll find some times when it's safe to send in your pet to get a few FI procs off, but it's usually going to happen that a doomfire's eventually going to kill him.

You can often remove the curse with an improved mend (especially come 2.4), and depending on your decursers, they may get your pet as well. I can't see why using a wind serpent would be that much more beneficial (I use one), unless it's for the 'fly in, lb, fly out' behavior, but even then your pet's most likely going to catch a doomfire eventually.

It's annoying to have a fight where you can't make use of your pet, but then again, Archimonde isn't a DPS fight, it's a survival fight.

I think we are seeing the effects of KC on a very tight rotation. In testing I achieved higher shots/sec with a 1:1 using the 2.7 spd.


Some of what we "know" may be based on incorrect assumptions (i.e. no interaction between KC and AS). When I have done Dr. Boom testing the 3:2 macro achieves higher shot output but this is not real testing as there is no pet involved.

From all the testing this is what seems to be happening.



In a clean 3:2 cycle the first auto shot is fired almost instantly after the first steady shot. The second auto is fired just before the start of the third steady. Any KC proc seems to bump the auto to the point of clipping into the next steady thus killing it.

That was also the highest shots/sec using KC of any test but at a 5.2 : 2 ratio it is a mana beast. I went through my mana bar 10 times on that test.

If 2.8 is the sweet spot that will make the [Crossbow of Relentless Strikes] that much nicer.

For reference: This does not only affect macros, it is an actual in game glitch. I do a manual rotation 100% of the time (thinking about giving a macro a shot, but I'm strangely opposed to the idea of more or less reducing myself to a one button spam)
This bug makes me CURSE the day I was born. It happens so often on kill command its amazing. It can however happen on ANY ability. I've had it hang me on popping rapid shot (hitting the button and being forced to move my character before the server registers that I've popped it seems to trigger this). Blizzard needs to figure out what is causing this and fix it, it may be because of the newer server side changes to abilities.. but I certainly hope not, as I live in Hawaii and my average latency on a good raid day is 200-250ms+

This is probably why I see it so often.. if I try to kill command at say 2% on a trash mob, the mob often dies before the KC registers and goes off, leaving it in the perpetual casting state and completely screwing my dps until I can get a arcane/multi/steady shot crit and use kill command again to reset it.

Extremely annoying.

I see the same thing. 3:2 runs smooth without KC but KC seems to push autoshot back even more. Someday they will unlink our specials from autoshot without messing it up like last time. What i dont understand is why noone at Blizzard can firgure out most of our issues in shot rotation is because of the .5/sec hidden cast of autoshot.

I've also noticed that the bug "resets" itself after a while. For example, last night I was doing a raid and the bug occured. So I go back to manually doing SS and KC (which works for me). But on clearing the way to the next boss all of a sudden the macro is working again.

The macro error doesn't reset after a successful manual KC. I tried that first. The macro doesn't reset after a /reloadui. I can't think of any rhyme or reason for it, unless it's just a time related reset. Also logging out and in to WoW has always fixed the bug for me, that seems pretty obvious.

Has anyone figured this out?

BTW, I'm using the standard 1:1 macro with WSR. I'm still trying to figure out the 3:2 macro and haven't grasped a complete understanding of it yet. Reviewing this forum to figure it out.

What macro are you using exactly? I fired 8,000 with each of the following two macros and never had the auto shot bug happen.

3:2
#showtooltip
/console Sound_EnableSFX 0
/cast !Auto shot
/cast Steady Shot
/cast [target=pettarget, exists] Kill Command
/script UIErrorsFrame:Clear()
/console Sound_EnableSFX 1

1:1
#showtooltip
/castsequence reset=1 !Auto Shot, Steady Shot
/cast [target=pettarget, exists] Kill Command
/script UIErrorsFrame:Clear()

Many people have reported that moving the KC to after the shots helped with the problem.

I think it only happens when your target dies. Something your 8,000 shots didn't run into.

Hmm, I have not run into it in any normal runs either however. Possible that the [target=pettarget, exists] evaluation avoids the problem?

Maybe. I've given up on the 3:2 macros so I've kinda lost interest in testing them any further.

This is the macro I'm using, althought I think my reset is longer and there isn't the ! before Autoshot. I do know the [target=petattack, exists] part is in there and that the KC line is after the castsequence. I double checked two nights ago.

BTW, simple question I should probably be able to look up, but google doesn't seem to be any help. What does the ! do?

I'd antidotally agree with this. (I know I shouldn't do that.)

The autoshot 'sticking' will happen whenever you can't finish a cast on your intended target as best I can tell. The one i had the most fun with was on the PTR actually. I was trying to squeeze a quick arcane shot off before I got ported out of the demon realm on Kalecgos but didn't quite get it finished. Of course this stuck auto until I was able to catch a portal again. I haven't been able to figure out anything besides relogging to fix the issue reliably as well.

Is the main difference between these 2 macro's mana consumption?

More mana/sec but higher damage/sec, yes.

There's nothing here that the theorycrafters on this thread haven't already found, but I'm seeing a lot of people asking the same questions, which indicates to me that conceptually the differences in rotations still has many confused. Here is some data (not perfect, but good) that may help clear things up.

There are so many factors involved (haste rating, armor pen, crit rate, and just plain serendipity) that keep rotation comparison from being an exact science. But after a day of testing, here is what I have to show, and the deductions I've made:

Tests were done with no haste effects (which with GCD as a factor with Rapid Fire, Quick Shots, Heroism), as clipping will occur at any time a normal rotation speed falls below 1.5 seconds. The tests are 2 full minutes of data.



Upon first inspection, the conclusion that is probably made will be "well total Damage and DPS for the second rotation is higher, therefore it must be the better rotation". Yes, and no. As I've already stated, the fact that in 2 minutes there is no foolproof way to control crit chance (and there are blocks) the total damage numbers are less useful than we'd hope. The more interesting data lies at the bottom of each test. First we see that steady shot does roughly the same ammount of damage (min/max) as auto shot. This is of course a conditional factor; steady and auto scale from RAP in completely different ways:

SteadyShotDamage = 150 + (WeaponDamage/WeaponSpeed)*2.8 + 0.2*RAP
AutoShotDamage = AmmoDPS*WeaponSpeed + (RAP/14)*WeaponSpeed + WeaponDamage + ScopeDamage

Also take into consideration that this is WITHOUT the 4 piece T6 hunter bonus (10% dmg to steady).

Regardless, the deciding factor at almost any gear level (completely excluding pets) must be total shots over the timeframe.

We can see that while there are 58 auto shots on the first rotation, there are only 46 on the second. Since auto shot is on a static timer, this indicates shot clipping. However, the second rotation has overall dolled out 7 more total shots. Again, assuming that steady damage is greater than or equal to auto, our conclusions are:

More mana usage
Higher potential damage output
Increased opportunity to crit (Go for the Throat)

Upon retesting the second rotation, the following data was cataloged:



Is it significant that this time the same rotation, with the same timeframe yielded approximately 7k less total damage? No, because again there is no way to account for crits over such a short period of time, therefore this total will be slightly variable from test to test. If we look at the pertinent data once more, we see that both the total amount of shots fired and ratio of auto to steady are almost identical.

I don't know if it's been mentioned but wouldn't testing the 3:2 macro be more accurate on a mob like the Servant of Allistarj in Blasted lands. The mob never dies. You can send your pet in and let it build 50k threat or so then reset your meter or start your combatlog and go all out on it. It should give you more accurate KC proc data.

The plot thickens...

Here are the variables:

-Different mob, lower level, lower armor (therefore, again total damage/dps should be discounted)
-Cheap arrows
-2 minutes of data (the duration of an IMPROVED hunter's mark)
-Immediate use of bestial wrath with rapid fire upon engage
-Pet is a windserpent, bite priority over lightning breath





The initial rapid fire for the 2:3 of course set off massive clipping, which explains the slightly skewed final ratio of SS:AS, and the fact that total shots fell behind.

The pet did quite a bit more damage on the 1:1 than the 2:3, but then again, in completely random fashion, the first had 31 crits to the second's 24 (pretty big gap, no way to account for it).

It still seems like SS will normally do more damage (again even without t6 4-piece), so without the sloppy clipping of a rapid fire or a heroism, I still feel that 2:3 will come out on top. The conventional knowledge is to swap to the 1:1 during stretches of haste.

Anyone see anything here I don't? Keep in mind that 2 minutes is not enough time to weed most of the chaotic factors, but I'm fairly sure there isn't enough gear in the game to sustain a significantly longer solo rotation.