This is an attempt to determine an MM "rotation" (in the loose sense of the term) based on the 7.1.5 proposed changes, and to determine whether any such rotation is both physically possible, viable in real-world play, and actually worth doing from a damage standpoint. I'm sure this is better suited to the Hunter Discord, but I don't have access to Discord from my current location, so it gets to go here.

Let's get one of thing out of the way first. I know the 7.1.5 PTR is "first-pass". Please don't respond to tell me that; you're not being clever. Blizzard has a long history of letting first-pass class changes go live, despite community complaints. It is better to be prepared than to have to adapt to this on the live servers without preparation. Worst case, I wasted some modeling time.

Second, I am not prepared to produce SimCraft values or firm dps approximations. Aspects of what I believe to be the potential best behavior are challenging to model in SimCraft. Consider this a preliminary analysis, mostly to check my assumptions.

Talents

The key aspects of the 7.1.5 MM hunter changes are the baking-in of the current Patient Sniper (mostly, sort of), the introduction of a new Patient Sniper with very different mechanics, and the brutal curbstomping of Sidewinders in any realistic scenario. This modeling exercise is intended to take advantage of the relatively large modifiers possible under the new Patient Sniper, while trying as best as possible to make up for the loss of Sidewinders. Therefore, the talents under consideration are:

• Tier 1: Lone Wolf, probably. Lone Wolf is the current baseline for MM hunters, so I'm assuming it for parity (again, I'm not actually prepared to produce DPS estimates yet). It is possible that Steady Focus will actually be worth consideration, because it minimizes the "build" phases, and eases the haste requirements necessary to handle the "burn" phases, but that adds several additional levels of modeling complexity, including modeling pet behavior and dps (which BM players will tell you isn't ideal at the moment). I suspect that the 18% gain from Lone Wolf will exceed the benefit of a shorter build phase, but the latter may make gear coefficients more obtainable. YMMV for now.
• Tier 2: Lock and Load. Because the 7.1.5 MM is based on maximizing the effectiveness of Aimed Shot, this is the only remotely viable option.
• Tier 3: Irrelevant here.
• Tier 4: Patient Sniper. The new PS allows extremely high burst damage windows in the proposed "burn" phases. These exceed any possible gain from the revised Sentinel except in pure AE situations (and even then, its CD is prohibitive).
• Tier 5: Irrelevant here.
• Tier 6: Barrage, probably. Because Barrage loses its focused cleave effectiveness in 7.1.5, it is relegated to use as a pure AE ability. However, the build/burn cycles necessary for ST dps in 7.1.5 are very strict, and my initial estimates are that AMOC is not a beneficial choice in pure single target situations. So you might as well have an AE tool for AE. Volley is always the wrong choice because it is functionally Focus anti-regen, and thus interferes with the core mechanics.
• Tier 7: Trick Shot. Sidewinders is dead. Piercing Shot is a large amount of damage, and probably superior in AE situations if targeted ideally, but is utterly unusable otherwise because of its focus-clearing mechanic. It doesn't come remotely close to the damage dealt by a burn cycle otherwise. Trick Shot remains terrible in 7.1.5, but is the best of a bad lot.

Build/burn gameplay

The proposal is thus: Determine the conditions necessary for optimal management of the new Patient Sniper, stacking large modifiers to multiple Aimed Shots in the 7 second window. With the nerf to Sidewinders, there are now only two ways to open the window: with an auto-shot proc followed by a focus-builder, or with Windburst.

In order to do that, a burn/build cycle of gameplay develops, where we expend our full Focus pool in order to maximize Patient Sniper use, and then enter a recovery period where we concentrate solely on refilling focus. An entire cycle requires approximately 14 seconds. This does not account for "traditional" burn buffs, like Heroism or Trueshot; those need to be modeled separately. This is simply what you do most of the time.

It is possible to fire 3 Aimed Shots in a Vulnerable/Patient Sniper window. The 7 second window opens when you activate Marked Shot (after the appropriate procs) or Windburst. The required GCD is tolled against the time, followed by 3 casts of a base 2-second Aimed Shot. The physical cast time requirements to fit that into a 7 second block are not challenging: [1.5 2 2 2]/h < 7 for any haste above 7.2%. However, that is insufficent to generate the necessary focus to perform the rotation. Modeling the focus requirement is nontrivial because focus regen occurs at second increments, which do not necessarily align with the start and end of activity. A post at mmo-champion, for example, suggests that, starting at 120 focus, you are guaranteed to re-cap focus during the casting of the first Aimed Shot, but that is not necessarily true, especially as increasing haste reduces the cast time. Regardless, any reasonable haste values provide enough Focus for MS - AiS - AiS, beginning with a full pool. In order to have suffcient focus for the third AiS in window, the MM hunter will need to wait for a regen tick following the 2nd AiS, while ensuring that sufficient time remains in the 7 second window to complete the casting before that window closes. At 20% Haste, that delay is approximately 0.4s, and will permit the completion of the final AiS 6.7s into the 7s window, assuming perfect gameplay and no lag. As the next situation will make clear, real-world use will require higher haste values and permit a safer margin.

Lock and Load procs, either banked from before the window opening, or triggered prior to the delay period, permit 2 free AiS that trigger the GCD but do not have a cast time. LNL permits 4 AiS in a window under ideal circumstances, and should be the goal of the 7.1.5 MM hunter. Focus management is not as strenuous here, because as we already determined, paying for 2 AiS in a window is trivial with max starting Focus. Fitting all 4 shots in before the window closes is a different matter. In order to fit 3 GCDs (MS, 2x LNL) and 2 cast AiS into the 7 second window, (8.5/h)<7 must be true. A calculator will tell you that occurs at 21.5% haste; however, at that value, the final AiS resolves with 0.004 seconds of Vulnerable remaining; clearly, this is not adequate. 23.2% Haste is necessary to allow for 0.1s of grace period on the debuff, and should be considered the lowest Haste for which this playstyle is remotely viable. This means that the 7.1.5 MM Hunter must value haste over other stats, and will require substantial re-itemization for most current MM players.

Note that these burn windows (especially the 3x AiS without LNL) essentially require that focus be hardcapped at 120 in order to execute correctly. At the suggested 23.2% Haste, for example, the delayed focus regen tick for the final AiS will fall short of the starting focus pool was even as high as 118! Using Windburst as the opener allows slightly lower starting focus pools (109 at the recommended minimum haste--yay?). The LNL-assisted burn window can be executed with as little as 81 Focus.

Burn windows essentially set the Focus pool to 0 (less so with LNL, but still close). In pure single-target environments, the fastest way to regenerate focus is to spam Arcane Shot. AS generates 5 focus natively. Here's the really bad news. Six consecutive Arcane Shots generates 30 Focus from the actual ability, and requires 7.3s, yielding an additional 86 Focus from passive regen at the recommended 23.2% minimum Haste. That leaves the hunter with 116 Focus in pool, which is not sufficient for a 3x AiS burn window out of Marked Shot.

However, in real-world situation, 5-6 Arcane Shots will suffice. Five AS require just over 6s of cycle time, and permit the LNL-assisted burn window, if it is off cooldown. Six AS will cap Focus if the previous burn phase was either opened by Windburst OR assisted by LNL. Additionally, the artifact talent Critical Focus causes AS crits to generate 5 additional Focus; if any of the 6 AS crits, then even back-to-back 3x AiS windows are possible. Because the entire burn/build cycle time takes 14 seconds, it is reasonable to assume that the procs to start the Marked Shot cycle are available (and, if not, ideally Windburst will be off cooldown).

In AE situations with 3 targets, MultiShot generates focus more efficiently than AS (although Multi does not benefit from Critical Focus). Note that if you do not expect the AE targets to live very long, it is probably better to Barrage, then build to a trimmed version of the rotation that does not attempt to maximize AiS count. Against durable AE packs, such as in M (or a few certain raid encounters), replacing AS with MS but preserving the burn windows is still the correct behavior, as Trick Shot will actually do nontrivial AE damage. The modifiers to AiS are quite large.

Does this actually do DPS?

Yes.

AiS is a hard-hitting ability, scaled to 250% weapon damage (times a coefficient chosen randomly per hit from [1, 1.038, 1.238, 1.619]; don't blame me, I didn't code this!). AiS is only used in Vulnerable windows, which increase damage by 100%. In pure ST situations, Trick Shot gets us another 15% (better than nothing, I suppose). And then we account for new Patient Sniper.

In the 3x AiS burn windows, those have Patient Sniper modifiers of 30%, 60%, 90%, respectively. In the LNL assisted windows, the exact modifiers depend on whether the LNL proc was banked before the window was opened, or whether it procced during the window. The simplest model is the former, in which the 4 AiS receive respective bonuses of 30%, 45%, 75%, 90%.

These are large multipliers to a large single target attack, and outweigh any benefit that could be gained from a shorter build phase or from interleaving other abilities like AMOC.

Can a human do this?

I think so.

Trueshot/heroism/etc notwithstanding, you open combat by activating autoattack. If you have been out of combat previously, you are likely at a full focus pool, and the proc mechanics will almost certainly give you an immediate Marking Targets proc to open a 3x AiS window. If the RNG hates you THAT much, you can open with a Windburst window instead.

After that, build using Multi or AS as appropriate for the target count. If LNL procs, open a window at your first opportunity if you have > 82 Focus. Otherwise, cap focus and open a window; the edge case of trying to hit 108 < Focus < 120 to open a burn window with Windburst is unlikely to occur often and frankly too much trouble to be worth calling out specifically.

In fights with movement requirements, you have freedom of movement throughout your entire build phase, and during the 2x LNL AiS in proc-assisted windows. The 3x AiS hardcast window cycle requires that you have around 5s of motionless gameplay (technically, there's a very short movement possible during the focus-regen delay before the third AiS, but that's on the order of 0.5s, and delaying too long is an immense dps loss). Try to open 3x AiS windows only when you can reasonable expect safety to stand still.

From here

I'd like others to check my math and ensure that my haste-manipulation break points are reasonable. I'd also like someone with more experience with SimCraft, if possible, to determine what the reasonable damage expectation is from this process. If not, I'll see what I can do about modeling at least the native case (cycling 3x AiS burn windows with 6x AS build windows) the old-fashioned way.

I expect this to remain a substantial dps loss in AE conditions (because of the lost interaction between Sidewinders, Vulnerable, and Marked Shot), and especially to "cleave" situations because of that nerf, compounded with the Barrage nerf. Ideally executed, my napkin math suggest that this may at least be reasonable comparable to current gameplay in pure ST dps, although there are a lot of assumptions there that may not hold true given more detailed modeling.

I decline to speculate as to whether this theoretical framework would prefer Mastery or Crit as its second secondary stat; clearly, gearing primarily for haste is required to hit 23.2% (or higher, for more safety padding). Mastery does not apply to the AS build windows; Crit increases the chance for Critical Focus procs that may shorten the window in some circumstances. On the other hand, Mastery is a larger, more reliable modifier to the AiS spam that is the meat of this proposal. I suspect that the delta between them will be smaller than it is on live, in any case.

And, of course, this is subject to change if the 7.1.5 math is updated.