Yes! it's really not a gain much at all. All it does it make chimera glyph work at odd haste values, which as a separate measure is a significant dps gain. But as you pointed out chimera glyph isn't always important.

I've been having a discussion on another forum and I can't seem to find a concrete answer so I'm hoping someone here would be able to help out. The discussion is focused upon weapon speed and how it factors into play even with weapon normalization. Below is what I assumed was the correct way to calculate it, although where I got my information from seems to be a bit outdated so it's left me questioning its correctness. Spoilered to save space:



So essentially I'm wondering if that's correct, and if not how would you calculate something like that? From how I calculated it weapon speed definitely plays a role in damage still, albeit a rather small one while top end damage has no effect.

Repins
In short, your spoiler is still correct, no changes in ranged weapons normalisation were instoduced since that time. Meanwhile it's somewhat incorrect to compare weapons looking at their speed only.
* Autoattack part depends on weapon dps only. The more the better, obviously.
* Weapon contribution in special shots depends on base weapon damage only. Speed is irrelevant as far as you have base weapon damage averages.

Still out of 2 weapons with same dps the one with lower speed is a bit better, because lower speed weapon will have higher base weapon damage. So for fixed weapon dps X and weapon speed difference 0.1 slower weapon will do arcane shot better by 1.15 (MM mastery)*1.12(AS glyph)*0.1*X, steady shot by 1.15*1.6*0.1*X*~0.713 (reduced by boss armor) etc. For Themios/Dragonheart the difference is about 90 damage fo AS, 92 damage for AiS, 58 damage for steady shot.

Can someone on the PTR checkout Rapid Recuperation please.

MMO Champions is reporting a channge to the effect that seems to indicate that the synergy with Rapid Killing - gaining 25/50 focus when gain it - is being lost. If this is really the case, this is a big loss to our AoE DPS as well as on trash and certain boss fights with adds.

Unless you were already at 1 second Aimed Shots without the haste trinket procs, the haste from the early double proc isn't wasted though - any speed up to your aimed hardcasts during careful aim is worth it, likely more so than deferring the procs till after RF wears off and risking losing a proc later in the fight.

To try to answer the question:

- the only way to time it properly would be to know exactly when you will be pulling, up to one icd ahead of time, which is difficult. What is doable though is just swapping hungerer out so it procs after MS, so for most people MS would then proc crit, while H would proc haste a bit later. But a crit proc during CA is also greatly devalued.

- the only way to know whether you'll lose a proc late in the fight is by knowing the fight duration, which you can at best only estimate.

The haste trinket procs at the start of fights is not actually wasted, in fact, it is pretty ideal (with the exception being if you have Bloodlust at the start of the fight as well). This is illustrated well in both the MM cases in the Best Possible DPS threads, which are extreme haste cases. And this is true even if you are a Troll and use Berserking to start too.

We do our most damage during the fight during the CA phase, where our number one damage shot is AI. The additional haste from the trinket procs has several benefits during the CA phase:
- Reduce AI cast time so you can perform more of them during the CA phase. Even with the extreme static haste amounts in the Best Possible DPS cases, glyphed RF, and both haste trinket procs during the CA phase, AI cast time is still above 1s at 1.007s. If you are a Troll, the AI cast time may be down to about 0.84s briefly, but that means that you can queue them better. If you have less than the static haste in these cases, then your AI cast times do not get as low and the trinket procs are even more beneficial.
- Increased autoshot frequency so that you can still do a lot of autoshots despite the lockouts from the AI casts. This increases the chances for a T12 4-set proc.
- Increased focus regen helping to power more AI casts with less SSs needed.
- Using the trinkets earlier will generally increase their uptime during the duration of the fight.

The only waste in regards to that haste is that your SS casts are already below the GCD soft cap even without the trinket procs, but that is really not a big concern since you should not be performing too many SSs – just enough to maintain ISS and provide enough focus for AIs. In the MM AI Best DPS case during the CA phase performs, 16 AIs and only 13 SSs are performed. That’s intensive DPS. If you take away the haste of the trinkets (with still including Berserking), you lose at least one AI and a SS for a sizeable DPS loss that cannot be made up by crit or mastery procs instead.

Also, if you do not have the trinkets on in combat, you are also losing a lot of DPS from the lack of the static agility (+900 including the 10% buff), equating to a large loses in AP (about 2000+) and crit (about 2.8%+). However, I assume you put the trinkets back on some time before combat starts.

Thus, save yourself the trouble and the DPS loss and just keep the trinkets on at all times. The haste being wasted on fast SSs is overcome by the haste’s benefit to allowing more AI casts during the CA phase when they are pretty much guaranteed crits. The DPS benefit of having the haste trinkets proc during the Standard phase when you would otherwise be unhasted is not nearly as beneficial as during the CA phase.

"Cetlysm gains 50 Focus from Cetlysm's Rapid Recuperation"
I also noticed the rapid recuperation is a buff now, it is 12 focus per 3 sec still, but its a separate buff from rapid fire(same icon as talent).
Also they fixed the 0.5 cast bug

To be clear, here is how an SS cast is 1.4s with 12.93% haste from gear and Pathing.

2s / 1.1 (10% haste raid buff) / 1.15 (ISS) / 1.1293 = 1.40001

At your haste level of 1872, your haste from gear is 14.62% and your over all haste including Pathing is 18.06%.

Your SS cast should be at:

2 / 1.1 / 1.15/1.1806 = 1.34s

Your SS cast should be 1.473 without the 10% haste raid buff. As MM, its hard to practive your shots on the target dummy if you do not have a friend available with the 10% haste raid buff.

At near end game gear levels like yours, you should follow the high haste recommendation in the guide.

Whitefyst,
I've been digging in this for a fair bit now and over and over i come across the same discrepancy in theory presented here & FD vs the practical results on WoL for example.

Looking thru the highest dps MM hunters on WoL shows quite clearly that vast majority of them do not use high haste builds, but rather stick to either 12.93% or 5.4% and prioritize either mastery or crit over haste. Which is completely controversial to FD and doesn't really sit will with your suggestions of higher haste plateaus.

We do have to take in account that T12 is mostly a very movement heavy tier of raid, but even on fights where it is possible to minimize your movement's negative effects on DPS, it is still quite clear that haste is far from the best possible stat.

In the best possible DPS thread, i posted about latency affecting the instant cast shots and was told that adverse effects of latency can be mostly overcome with 'custom lag tolerance' option. Upon which i took the time and did some digging in the combat logs - even with the option enabled i can still see the GCDs being at approximately 1.3 seconds with 300 ping. Could anyone else check it in their logs and verify this? (currently using mousewheel to ensure the tightest possible casts)

If this is the case, would that alone not stand very strongly against the high levels of haste on premise that adding extra shots also adds your latency into the equation? In which case it would make a lot more sense to have less shots that hit harder via crit as well as more mastery procs.

Actually, if you pull most of the top 10 from Baleroc heroic 25man, they are High Haste rating:

Rank 3 Skunkzy - EU Nathrezime
Rank 4 Geoph - EU Nathrezime
Rank 9 Eiwo - EU Auchindoun

These characters are all over 1850 haste rating with 3/3 pathing included.

However, I don't think you are wrong. Whitefyst recommends forging into so much haste that both trinkets pop haste, yet I haven't found a single hunter that has done this. Especially one that has ranked highly in the only real dps fight that there is, Baleroc. It's all well and good to do amazing numbers on paper, but when your fighting a boss and you factor in the mechanics, the evidence stacks up against super high haste.

From my own personal experience, its difficult to judge what is better. High mastery is nice for fights like Beth'tilac where we are AoE'ing but on fights like Ragnaros, when you need massive amounts of single target dps to race the Meteor spawns, its seems better to have high haste.

It would be amazing if we could reforge on the move to really maximize each fight!

I am interested to see Whitefyst's comments on these two posts now =)

Considering the recent discussions on haste builds theoretical DPS versus mastery builds practical DPS, I thought that I should do some analysis on the situation using FD to roughly quantify it. For this analysis, I used my character instead of the BiS setup in the Best Possible Thread since I am more comfortable with the setup for my character. My ilevel is 385-6, so the result may not 100% translate to 391 gear set, but I assume that there will be a lot of parallels.

The haste build is the setup currently on my character where I have just enough haste rating to be barely above my crit rating so that Matrix Restabilizer will always proc haste. FD does a great job of modeling the haste proc anytime it occurs. The setup uses the RF, SS, and AS glyphs since I still cast AS when under no dynamic haste effects. For this setup my stats on gear are:

Stat	Amount
Agility	6009
AP	190
Hit Rating	968
Crit Rating	2093
Haste Rating	2100
Mastery Rating	654
WQ%	24.46%

For the mastery setup, I kept all talents, glyphs, and shot priorities the same and only reforged gear to get as close to the 1235 haste rating for the 12.93% haste case with 3/3 Pathing. For reforging, crit still had priority over mastery when I could reforge into either.

The main difference in this setup beyond the reforging is that Matrix Restabilizer no longer only procs haste. If Hungerer is not currently procced when Matrix procs, then Matrix procs crit instead of haste. FD does not model the crit or haste proc dynamic choice and instead assumes it is always a crit proc. FD does not accurately model the crit proc since it averages out over the whole fight. This actually hurts this case some in FD since the additional average crit from the proc at the start of the fight during the CA phase is mostly wasted since AI and SS will already be near or at the crit cap. Plus, in actual game play, Hungerer will usually proc first at the start of the fight making Matrix proc more useful haste during the CA phase.

The resulting stats on gear that changed are:

Stat	Amount	Delta
Crit Rating	2120	+27
Haste Rating	1358	-742
Mastery Rating	1369	+715
WQ%	32.84%	+8.38%

The comparison of these two setups in shown in the table below:

DPS	Haste Setup	Mastery Setup	Delta
Total	36558	36216	-342
Hunter	31752	31688	-64
Pet	4806	4528	-278

As can be seen, the total DPS loss is 342, which is the equivalent of replacing about 2.7 AIs with ASs, when moving. Considering that AI cast time should normally be longer than the AS GCD and costs more focus, you should be able to replace more ASs per AI than 1 for 1 when on the move. Hence, I can see how the mastery build can be comparable DPS on small movement fights and maybe better DPS on high movement fights. Also better modeling of Matrix dynamic haste or crit procs will add DPS to this case as well. Also need to account for the negative difference in autoshot DPS from the haste setup when having less AI lockouts.

Furthermore, the majority of the DPS loss is in pet DPS due to about 5.5% in WH uptime. Only about 64 DPS is lost in hunter DPS.

More details on the difference between these two cases are:
- The mastery set up case does more damage per shot since it has a higher crit rate due to the little more crit rating on gear and the average crit rating from Matrix procs.
- The shot differences between the two cases are 10 AIs and 5 SSs in the haste setup versus 10 ASs and 30 WQs. The number of autoshots are the same between the cases due to the AI lockouts.
- The 5.5% difference in WH uptime occurs despite the same number of shots that proc Sic’Em. The difference is more AIs during the CA phase with almost guaranteed Sic’Em procs in the haste setup. This difference would be reduced if FD modeled possible Matrix haste procs during the CA phase, which would also reduce the pet DPS differences.

The next item I investigated in regards to these setups is the impact to AoE DPS. Since mastery is touted as a big gain for AoE DPS, I figured that the mastery setup would be far superior. The results in FD contradicted this as shown below.

# Targets	Haste Case	Mastery Case	Delta DPS
1	25817	25888	71
2	31084	31099	14
3	36352	36309	-43
4	41619	41519	-100
5	46887	46729	-157
6	52154	51940	-214
7	57421	57150	-271
8	62689	62360	-329
9	67956	67571	-386
10	73224	72781	-443

The AoE cases used the same setups as for the single target cases but changed the abilities used to just ET, MS, SS, KS, and AI MMM procs. I also reduced the fight duration to 1 min with not including any RFs or BLs.

The mastery case actually has better single target DPS mostly due to the higher crit rating but also due to the extra WQs. However, surprisingly it has lower multitarget DPS. The reason for this is that the mastery setup did 2 less MSs due to less time for instances with longer SS casts to regain focus and due to the lower base focus regen rate at this haste rating resulting in less focus available for MSs.

One major factor not accounted for in this FD analysis is the benefit of Rapid Killing returning 50 focus per AoE killing blow. The bonus focus will favor the mastery setup case since it will help to eliminate required slower SSs to regain focus once AoE targets start dieing. Just reducing the MS cast difference by 1 due to RK adds another 230 DPS per target to the mastery case, which would make the mastery setup superior to the haste setup as expected. The mastery case would now do about +173 DPS per target over the haste case.