I agree with Sidis in that in general you architect your character for boss kills, even during progression.

But I also understand part of Nooska's point as well. If it was that the KS glyph was a clear, overwhelming favorite, then his discussion would be invalid. But with the choice being currently close, fuzzy, and basically preferrential, then part of his point is valid.

To illustrate the situation better, below is a tabular breakdown of the DPS numbers my analysis used over each of the fight phases for the case with using Bloodlust early:

	Duration	KS glyph	CS glyph	CS + Term
CA phase hasted	20.00%	16470	16627	16627
Standard phase	42.00%	9562	9656	9656
Standard phase hasted	13.00%	11782	12017	12017
Term phase	5.00%	9562	9656	10171
KS phase	20.00%	10614	10059	10917
Total	100.00%	11443	11438	11635
Difference from KS glyph			-5	+192

If the overall boss DPS of the 2 glyphs are equal and the choice truly preferrential, the difference is that the CS glyph provides a little more damage over the first 80% of the boss fight, helping a progression raid go a little deeper in boss health faster. On the other hand, the KS glyph allows a lot more DPS over the last 20% of the boss fight, helping the raid finish the boss off better once you are able to get it below 20%. Which is more important depends on you and your raid's preferrence. I personally prefer being able to finish the boss off faster, if all things are equal.

The part of Nooska's argument concerning if you do not get deep enough into the boss fight to utilize a talent or ability that its value is zero is an invalid argument. Its value is the same regardless of whether you get to the point where you would use it or not. Sure, its value will not show up in the DPS numbers for the wipe log, but that is just a side effect of the wipe. On the other hand, anytime you make changes to inflate your DPS numbers for a wipe, you are then hurting your overall DPS if your raid progresses farther and has a chance to kill the boss.

Edit: I incorrectly linked in the wrong DPS for the KS glyph in the table. I have corrected the value from 11640 to 10614. This decreased the total numbers for the KS glyph in this case with Bloodlust early from 11648 to 11443. Hence, the KS glyph is equivalent with the CS glyph without Termination and is about 200 DPS worse than the CS glyph with Termination. Note though that the with Termination numbers does not factor in the DPS loss from the ISS downtime, which probably evens things out. You are also giving up the benefits of alternate talents to Termination, like Bombardment.

Agreed, and this is one of the items I mentioned in the guide. The problem though is that there really is no good way to accurately determine what % the last 20% is without looking at a raid log, but then the amount will differ from raid to raid (some like to use Bloodlust early when everyone is alive) and from boss to boss and is affected by the number of DPS remaining alive and their available remaining attack resources. Furthermore, if you modify the KS phase duration, then the CA phase, which favors the CS glyph, probably needs to be modified as well. That is when the MM DPS is the highest due to CA. It is also where most DPS classes have guaranteed trinket uses and also use CDs. Then even during the standard phase, we are going to use our second RF-RD-RF and other classes will probably have DPS boosts as well, shortening its duration a little too from the steady state DPS rate.

Thus, I just used time percentages equal to health percentages as a rough estimate with making assumptions that the CA and KS shorter durations than listed roughly balance out.

Note that the numbers in the previous post were for the analysis with Bloodlust near the start of the fight. Here are the numbers with assuming Bloodlust at the end, which favors the CS glyph case a little more since it provides more focus to use the maximum Termination DPS rotations more instead of using focus regen rotations.

	Duration	KS glyph	CS glyph	CS + Term
CA phase hasted	10.00%	16470	16627	16627
CA phase unhasted	10.00%	12357	12599	12599
Standard phase unhasted	45.00%	9562	9656	9656
Standard phase hasted	10.00%	11782	12017	12017
Term phase	5.00%	9562	9656	10171
KS phase hasted	13.33%	12667	12569	12800
KS phase unhasted	6.67%	10614	10059	10917
Total	100.00%	11240	11299	11412
Difference from KS glyph			+59	+172

Edit: After looking over this post, I realized that I had linked in the wrong numbers in a couple of places in these tables which affect the analyses. The KS unhasted phase number for the KS glyph is 10614 instead of 11640. In addition, I still had the unhasted numbers in the KS phase hasted row for all columns. Hence, the KS glyph is a little behind the CS glyph without Termination and is about 172 DPS worse than the CS glyph with Termination. Note though that the with Termination numbers does not factor in the DPS loss from the ISS downtime, which probably evens things out. Either way, the differences are close enough currently where the choice is prefferential.

Edit2: Also realized that for the CA phase unhasted numbers I had used the numbers for the rotations without AS instead of the numbers for the rotations using AS, which are better unhasted due to the longer SS cast times. Updated Edit 1 as well as the table. The guide will be updated later today.

Edit3: Okay, found another problem. Sorry, these tables were new to the analysis this time and were added at the end when tired plus some characteristics changed from the last analysis. For the unhasted CA numbers, the KS glyph is slightly better at the higher crit rates now to use the special CA rotation during the CA phase when not hasted; however, the CS glyph cases are better off using the standard rotations during the CA phase when not hasted since 1.5s SS cast time results in 0.5s deadtime when just casting SSs between CSs.

It seem we've been buffed big time

Aimed shot's base damage is doubled. Chimera is up by 20%. Another 20% buff was certainly meant for BM, but the side effect for us is the increased value of Resistance is Futile. Overall it seems like a very significant damage increase.

PS. That is PTR data so far, so don't expect to see it on live yet. I just realised I didn't make it clear originally.

Jan. 12 patch notes have some very interesting items for us, in particular:

Some of this we already knew, some is unclear if it's in addition to or supercedes the previous changes (ie, Chimera Shot damage), but all welcome news. I'm particularly amazed by AutoShot on the move, after years of Blizzard insisting that would be catastrophic. Perhaps this is a part of the philosopy shift that saw us become a full-fledged physical class? In any case, it certainly increases the value of our Artisan Quiver mastery, especially as so much movement is required now.

Further, by virtue of increased shots, it should also offer a boost to Wild Quiver and therefore the value of mastery rating.

I don't know how much value you should put on the racial values. Blood Elves are ranked lowest with Tauren, despite having a better racial than Tauren, Undead, Humans, Night Elves and naturally Dwarves with no gun. Also, their base Agility is higher than all of them, save the Night Elves.
Also they are resistant to Arcane rather than magic on a whole.

My current understanding is that the shots are only similar. Neither requires any action to fire and both are based on weapon damage with doing similar damage. I believe that WQ is unaffected by Artisan Quiver since it is not an autoshot, but we had thought that it was affected by it at one time.

WQ definitely does not provide focus from Go for the Throat. That would be a little overpowered in AoE situations with all of the WQ procs.

Ok, playing with some numbers. Dropped into FD and had a DPS of 13.6k in MM (to give you an idea of gear I am basing this on)

Assumptions:
Arcane hits for 7336 (from FD)
Aimed hits for 12082 (from FD)
Buffs put Arcane at 8436 and Aimed at around 13290 (I guessed the weapon damage buff would buff aimed by about 10%, I honestly have no idea)
Aimed's cast time is 2.4 seconds


I wanted to see at what point casting Aimed Shots would be better than Arcanes.
Aimed gets the benefit of Piercing shots, so for an average aimed, add on 30% * Crit rate (38% in my setup)
Aimed also gets the benefit of Careful Aim, which increases it's crit by 60% for the first 20% (which is why the crit rate is higher for aimed than for arcane). That also means your pet is doing more damage from Sic'em.
Aimed is a 50 focus cast, Arcane is 22, so that has to be figured in at some point too.


But for raw numbers...
Average Aimed damage = Shot damage + Piercing damage = 13290 + 13290 * .3 * .38 = 14,819
Average Arcane damage = 8436

So to be equal DPS, aimed would have to have a cast time of 14819 * 1 / 8436 = 1.76 seconds
That means your haste has to be 1 / (1.76 / 2.4) = 36.4%

So at 36.4% haste, without considering Focus, Arcane and Aimed do the same dps, so you can interchange them based on focus requirements, timing, etc.

Given the higher focus costs, it is probably only worthwile if you are under hasted effects (bloodlust, rapid fire), but at that point, Aimed would be a much better dps choice

FD has PTR changes enabled as a setting. The difference between aimed and arcane is far more than you posted.

For my gear (full 359), Aimed with Piercing Shots is 26K and Arcane Shot is 13K. We have 40% + attack speed increase without cooldowns with ISS.

Now, the focus cost can't be ignored. My napkin math says we'll end up using Aimed during Careful Aim, Rapid Fire, Bloodlust and Termination range (because of the additional focus gain).

Considering that Aimed Shot is also affected by Careful Aim (and Piercing Shots), it looks like MM hunters should no longer spam Steady Shot for the first 20% of the fight, and instead use Aimed Shot to bleed off extra focus.

Also: The patch notes said the weapon damage of Aimed Shot was raised from 150% to 200%, but the Wowhead currently lists 95% for live and 65% for PTR. What gives?

Aimed Shot on PTR has 200% weapon damage for the casted version and 102% for the MM Proc. I have no idea where blizzard pulled the 150% number from. As for WoWhead values, they might be showing the lowbie weapon damage, though things have been sorta bugged once they introduced the weapon damage scaling to limit lowbie dps on WoWhead tooltips.

With Aimed getting a lower cast time, would this increase haste's value some or will it still be the same due to the MM Proc once the bug is fixed and we'll still aim for certain plateaus? Am I safe in assuming that Crit > Mastery > Haste is still the way for MM reforging? (Haven't looked into MM theorycrafting all that much given how poor the spec has been doing for some time)

Yes, as long as Aimed Shot cast does not reset the autoshot timer anymore (which is a big wish), there is a lot of potential for using the cast AI in the MM rotation. Of course, even if AI works out better than AS theoretically, keep in mind that a long cast attack is less beneficial on movement fights or fights with a lot of interruptions to casts. Hence, under those conditions, you would want to use AS instead even if it is not quite as good in the theoretical situation.

Here is the current PTR shot data for my character:

KS: 17083
CS: 39161
AI: 23766 (38453 during CA)
AS: 12202
SS: 7969 (13102 during CA)

From my traditional MM analysis in the guide and not factoring in the cast AI:
1) Definitely still want to ignore AS during the CA phase as long as you are under large dynamic haste effects (RF or Bloodlust) where your SS cast time is close to 1s. You still want to use AS in the CA phase under slower SS cast times. Hence, nothing has changed in that regards.
2) Although KS got buffed, these changes do slightly favor the CS glyph, especially with Termination. There are still arguments and certain fights where the KS glyph is better, but the scale is leaning towards the CS glyph.

I have not updated my shot spreadsheet analysis for checking out the shorter cast AI (that will take a bit of time), but here is some quick napkin math. With the minimum 5.4% combined haste from gear and Pathing, ISS, and the 10% raid buff:

- During RF, we are at 86.6% haste and would have an AI cast of 1.2857s.
- During BL, we are at 73.3% haste and would have an AI cast of 1.3846s.
- With no dynamic haste effects, we are at 33.3% haste and would have an AI cast of 1.8000s.

Scaling the damage to damage per second of cast:

- During RF, AI = 18485 (29909 during CA)
- During BL, AI = 17165 (27772 during CA)
- Unhasted, AI = 13203 (21363 during CA)

As can be seen, all three AI numbers are greater than AS damage at a cost of 2.27 times the focus. Furthermore, although AI triggers Sic'Em too, you can do more than 2 ASs per AI as far as focus goes for more chances for Sic'Em procs.

Rough conclusions per phase:

CA phase:
- If under dynamic haste effects during the CA phase, the AI cast is pretty low and does significantly more damage per cast time than AS or SS. Since the special CA rotation while hasted generates well over 100 extra focus for the KS glyph case and a little over 100 focus in the CS glyph case, it is safe to state that casting 1 to 2 AIs (depending on focus)during the CA phase when hasted would be a DPS gain. I state 1 to 2 instead of 2 since we are not only giving up the focus generation from the SSs being replaced by the AIs, but we will also have less frequent free AI casts from MMM.
- If unhasted, the answer is a little trickier and I will need to wait until I fully model it to provide a definitive answer, but since the difference between 1 AI versus 2 AS is about 14K damage and 0.2s versus 6 focus and twice as much chance to proc Sic'Em, the data would seem to indicate that using AI instead of AS would still be beneficial during the CA phase when unhasted.

Standard Phase:
- When unhasted, the trade off between 1 cast AI versus 2 AS is 0.2s versus 638 damage, 6 focus, and twice as many chances for an Sic'Em proc. Thus, during the standard phase unhasted, it seems that AS is the winner, but not by much.
- When hasted, the situation seems to favor casting AI.

Term Phase:
The results are similar to the Standard Phase except that the additional focus supports the extra focus cost of AI. Need to model this before I can decide what type of rotation is best.

KS Phase:
Without Term, the results are similar to the Standard phase with the additional focus generated during the KS phase definitely supporting the extra focus cost of casting AI such that it is not a factor. With Term, I need to model it since the focus cost is definitely a factor when skipping the second pair of SSs for reduced ISS uptime.

I will not be updating the MM guide for any of the PTR changes until they become live.

Note that I have a concern with this data. As it looks right now and with how AI scales much better than AS, it appears that AI cast will marginalize the use of AS. Using AI cast is definitely superior during the CA phase and can be in the Term and KS phases depending on the situation. Even if AI cast is not the better choice in other phases, especially when not hasted, with scaling of gear, AI should eventually be favored.

Hence, either Blizzard wants us to normally use AI cast instead of AS instant (except when on the move) or I think that the AI and AS numbers will be adjusted some more so that AS is not marginalized and is still a key MM shot outside of the CA phase. Thus, I would not be surprised if AS damage is adjusted up some more and/or AI damage is reduced a little.

Furthermore, factoring in the AI cast will increase the value of haste for MM over the current range between the 5.4% min and 12.93% soft ceiling. Even at the 12.93% haste from gear and Pathing where the SS cast is 1s during RF, the AI cast is still 1.2s, so additional haste affects it. When under BL, 21.61% haste is needed from gear and Pathing to get SS to a 1s cast. In that much static haste, during a RF the AI cast is still over a GCD at 1.114s.