Now that FD models haste procs and the Max DPS thread is started, I am sure that I am going to get a lot of questions concerning the high haste situation seen in FD.

From what I can tell, these high haste cases are real. They result from the convergence of several factors:

1) The Troll racial Beserking. The 20% haste from this talent is enough to lower the AI cast time below the 1.8s guideline for use with only 6.1% haste from gear and Pathing.

2) The haste trinkets, especially the heroic versions. The heroic versions of these trinkets only require 12.2% haste from gear and Pathing to get the AI cast times below the 1.8s guideline. The normal versions of the trinkets only require 13.8% haste from gear and Pathing.

3) The T12 4-set bonus. The value of the T12 4-set bonus increases with the more higher damage, high focus costing shots that it can be applied to and allow.

Whether you chose to use AI as your focus dump all of the time or only when your AI cast time is below the 1.8s guideline, the result of the above 3 factors is that with a reasonable amount of haste on gear, you will have faster AI cast times more often resulting in casting more AIs and taking more advantage of the T12 4-set. Even the AS unhasted focus dump case has more AIs (60) cast than ASs (29).

With casting a higher ratio of AIs and with AI cast time being affected by haste, the benefit of haste becomes much higher since it reduces the AI cast time. It also reduces your static SS cast times so that you can generate focus faster for casting more AIs and allows for faster autoshots, resulting in more T12 proc chances.

Despite that fact that AI locks out autoshots, with how fast the autos and AIs are in this case, you do not lose many autoshots in the case as opposed to the standard case. Thus, you only lose at most 1 or 2 T12 4-set procs, but that is made up for by utilizing a higher percentage of the procs on AIs instead of ASs.

Although the value of this case is reduced a little if you are not a Troll since Beserking is key to maximizing the case in FD, but it is not the discriminator since it only has 20s of uptime in a 5 min fight for only about a 6.7% uptime. Using an Orc in FD is only a 400 DPS loss. This number is a little misleading for the following reasons:

a) The loss would actually be higher in FD if we had the control in FD to use the first Beserking later in the fight when we otherwise have no other dynamic haste effects. It is a waste to use it at the start of the fight on top of RF and both trinket procs. 0.835s AI and 0.536s SS cast times waste pretty much the whole benefit while waiting for the GCD. It would be much better to utilize it at a later time to allow more AI for AS replacements.

b) Blood Fury (and other similar racials) are still implemented as averages. Blood Fury has a 15% uptime in a 5 min fight, which works out to an average of 175 AP through the fight instead of 1169 AP during the 45s of its effect. I am sure that in this Max DPS case that having about 1000 extra AP during the CA phase with near 100% AI and SS crits with near or below GCD cast times on each and also during two trinket procs later instead of the 175 AP constant would be a big DPS boost for Orcs.

The big discriminator in this case are the two trinkets. Even with overlaps in their procs, they had near 50% uptime combined (140s of 300s = 47%). Although some of this time overlaps RFs and BL, it is still more AI for AS replacements. Even when overlapping RF and BL, the increased AI cast time and autoshot frequency is very powerful. The two keys with these trinkets are to have enough static haste so that the Matrix Restablizer will always proc haste (meaning must have more haste on gear than crit) and so that your AI cast time sufficiently low enough when either trinket procs to make it worthwhile to use instead of AS.

The benefits of both trinket procs on top of RF during the CA phase makes the AI cast time short enough where it is beneficial to totally forgo SrS and CS during the CA phase.

I do not plan to update the MM guide with this information yet since I first want to fully understand it. Second, I need to figure out how to incorporate it into the guide since it is not a static condition. It is a temporary condition dependent on haste trinkets and the T12 4-set.

I think we need to get over fixed rotations or haste plateaus while talking about 4pct12. We are now very procc dependant, both for the 4 piece bonus and the trinkets available this tier. Analizyng the max dps case, there´s hardly a cycle that repeates itself within the next 7-10 cycles, even when there is only 3 different shots. There are cases where CS is used on CD, and others where it is delayed up to 4 secs. Its time to move and start thinking fully into priorities and not try to fix every cycle, ss we are hardly gonna repeat the same exact cycle very few times in the same fight, even less to see it take place enough times to set a suitable haste plateau.

To be clear on a couple points:

1) Concerning an AS focus dump cycle. Since AS is so much less damage than CS, it is important to have tight or somewhat tight CS cycles. Delaying your CS cast too long is a DPS loss. Next, the rotations I list in the guide are not fixed rotations, they are guidelines for maximizing DPS and are dynamic. I do list a base cycle that is performed on cycles without procs (MMM AI or T12 4-set), but these are developed with factoring in the impacts of the procs and change from cycle to cycle depending on which procs are present and your current focus level. They also have flexible order within the constraints of maintaining ISS and focus boundaries. Even with factoring in haste trinket procs, you still perform the same base cycle with just casting extra ASs or SSs with the extra cycle time depending on your current focus levels. Cycles should be highly repetitive.

2) Concerning an AI focus dump cycle, as I tried to indicate in the previous post, having a fixed rotation or even a base rotation really does not make much sense. There is less reason to cast CS off CD. With AIs long cast time and high focus costs, in order to maximize DPS you often have to squeeze in extra AIs on one cycle and then spend later cycles casting extra SSs instead to balance the focus. Plus, changes in dynamic haste levels affect AI cycles much more than they affect AS cycles. It is the current common use of haste trinkets that make designing a base AI focus dump cycle futile than impacts of the T12 4-set bonus does.

3) The variation in CS CD usage is a real situation when using AI as your focus dump as indicated in the previous item; however, the variation in FD is not necessarily ideal. Sometimes FD makes decent but not optimal decisions on which shots to perform because of its rules based methodology. For the AS focus dump case, there should ideally be very little variation in CS CD usage. It should always be used within 1s of CS coming off CD. However, FD’s rules can result in CS being cast later in an AS focus dump case, which is a DPS loss. So although FD and other simulations are great tools, you cannot take the information that they provide too literally as gospel.

Totally agree with all the 3 points. However, I was talking about the AiS focus dump. I don´t think the AS focus dump will be competitive unless we are talking about a massive movement fight. Yes, the max dps case for the AS focus dump setup is only 400 dps behind, but thats only beacause it is not really dumping focus with AS, as AS represents 4% of the total DPS, half of what it should be compared to t11 numbers. It is simply an AiS focus dump setup that uses AS when completely unhasted

Well, what I don't understand is how haste can be valued good in the Arcane dump rotation, it makes no sense with the set rotation we basically have and already being way focus posetive because of T12 set. I find myself sometimes just spamming Arcane Shots just to not cap. How can haste be valued good when you can't even use the focus?

I agree that the Best Possible DPS in such a such tool cases should always be taken with a grain of salt. Part of the reason is it is an ideal fight that is being simulated, and no fight, even a Patchwerk fight, is every 100% ideally performed. Second is the modeling of the tool can result in not so practical of results. A third is that it requires the ideal gear set that the vast majority of players never acquire.

The value of such cases though is that it gets players trying different things in the simulations to attempt to increase DPS that they may not have ever dreamed of trying in game. Some of these changes are not practical and only are beneficial in the simulation for a particular case. However, some of the attempted changes are viable options for end game playing or provide insight into ways to improve play.

Anyway, any change that allows us to reasonably perform more AIs instead of ASs is going to increase DPS, but that indeed has to be balanced against the situation where fight mechanics may reduce the opportunity to perform more AIs and require a decent number of ASs to be cast. That is why I am personally not a fan of the AI hardcast only all the time optimization. I try to optimize around a decent number of both AIs and ASs being performed.

For my current gear, which is a decent amount below current BiS (376 ilevel), this recent whole max DPS case has not affected my playing approach at all. With only one haste trinket and with it being a non-heroic version, it neither gets my AI cast time down to a low enough cast time at reasonable haste amounts nor has sufficient uptime for me to make any changes.

The fact that some of the gear in the cases in the thread may not exist does not change the fact of the benefit illustrated by these cases of having enough static haste to ensure that Matrix Restablizer procs haste and being able to utilize both the haste trinket procs in order to reduce AI cast time sufficiently during the procs to increase the number of AIs potentially performed during a fight.

To illustrate, I took the AS focus dump case from that thread as a base and then modified it to bring it down closer to the current recommendations for an AS focus dump case in the MM guide. I also changed the CS glyph to the AS glyph. The table below summarizes the changes in stats.

Stats from Gear	Max DPS	Standard	Delta
Agility	6341	6341	0
Stamina	7000	7000	0
Attack Power	190	190	0
Hit Rating	961	963	2
Crit Rating	1876	2212	336
Haste Rating	2589	1559	-1030
Mastery Rating	640	1332	692

As can be seen, agility and AP were unchanged in the case. The only difference between the two runs is trading off 1030 haste rating versus 692 mastery and 336 crit. This is basically losing about 8% haste from gear, but after being multiplied by other haste effects it is essentially about 40% haste. The changes in crit rating plus the FD modeling of the Matrix Restablizer proc now being crit always averaged over the fight, results in about +4.6% crit.

So the next question is how much did this change affect our average shot damage. The table below shows this. The table below includes the benefits of PS to SS, CS, and AI. It does not include the buff to crit on SS and AI during the CA phase.

Shot	Max DPS	Standard	Delta	% Delta
Serpent Sting	3324	3396	71	2.14%
Kill Shot	31644	32748	1104	3.49%
Chimera Shot	53527	56178	2651	4.95%
Arcane Shot	14575	16893	2318	15.90%
Aimed Shot	56587	59390	2803	4.95%
Steady Shot	12489	13108	619	4.95%
Auto Shot	10066	10417	351	3.49%
Proc DPS	5269	5281	12	0.23%
Wild Quiver	9624	9959	336	3.49%
Flaming Arrow	9763	10103	341	3.49%

As can be see the change in damage is not big but still significant. Each SrS tick does about 2.14% more damage. The three PS affected shots each do about 5% more damage. AS did about 16% more damage due to the glyph plus the additional crit rate. All other shots do about 3.5% more damage. The question then is does the chage in shot totals make up enough for the differences. The table below shows the change in shot totals.

Shot	Max DPS	Standard	Delta
Serpent Sting	1	1	0
Kill Shot	5	5	0
Chimera Shot	29	27	2
Arcane Shot	29	52	-23
Aimed Shot	61	42	19
Steady Shot	142	135	7
Auto Shot	125	132	-7
Proc DPS	1	1	0
Wild Quiver	95	128	-33
Flaming Arrow	14	14	0

The 2 difference in number of CSs is due using the CS glyph. Besides the 2 additional CSs, the higher haste case gained 19 AIs and 7 SS versus 23 more ASs. That is a significant amount more of DPS. This increase is partly due to The Hungerer procs being at high enough haste to meet the threshold to cast AI when under no other dynamic haste effect. It is also due to the Matrix Restablizer now always proccing haste and resulting in performing AIs when under no other dynamic haste effects and making the AIs under other dynamic haste effects cast faster.

Although the faster haste case had 7 more SSs, it was not enough in the simulation to get another Flaming Arrow, although an additional one or more could occur in game. The faster haste case did have 7 less autoshots due to the increased AIs and autoshot lock outs. This did result in 1 less T12 4-set proc, but with it and other procs being used on ASs, the net focus benefit from T12 4-set procs was higher in the faster haste case. The final shot difference is 33 less WQs.

So to what DPS difference does this equate? See the table below.

DPS	Max DPS	Standard	Delta	% Delta
Combined	40301.64	39526.12	776	1.92%
Hunter	35317.45	34845.59	472	1.34%
Pet	4984.19	4680.53	304	6.09%

The faster haste case is about 2% better on DPS, with DPS being higher for both the hunter and the pet. Pet DPS is higher despite the lower crit rate due to the faster attack speed. Note that Sic’Em procs were about the same in both cases since there was a similar number of AIs plus ASs performed (90 and 94) with the more AIs performed in the CA phase evening things out. The less number of autoshots and lower crit rate did result in a little lower focus regen from GftT from the faster haste case.

Now that the case comparison is complete with the faster haste case winning, let us look at the practicality of the situation for actual game playing. Factors include:
- Faster SS cast times are better for game play favoring the faster haste case
- Things happen which prevent the ability to perform the rotation, delaying CS casts and favoring the lower haste case that does not have the CS glyph
- Faster AIs in general result in less chance for each to be interrupted or pushbacked favoring the faster haste case
- Having to move and being afraid of interruption and pushback boss abilities occurring can result in needing to perform more ASs over AIs. This hurts both cases, but obviously the higher haste case more since it performs more AIs and does not have the AS glyph.

As a compromise to the practical benefits and disadvantages of the higher haste case, what I recommend from a practical standpoint is to still use the AS glyph instead of the CS glyph. In the sim, this change is about a 360 DPS loss, which takes away a little under half the DPS benefit of the higher haste case. However, due to the practical situations affecting CS casts and having to perform more ASs than ideal, using the AS glyph instead of the CS glyph in the high haste case could actually be a DPS gain in practice.

With that change, the difference between the two cases is about 410 DPS. This is a significant amount at about 1% of the DPS. From practical standpoint that difference is smaller, say around 300 DPS, which is small enough where I say that the approach between the high and low haste cases with the haste trinkets is a personal preference for what works best for your play style.

If I understand the context of your comment correctly, the higher haste case is still valuable for the AS unhasted focus dump case since:
1) It allows you to use the much higher damaging AI as your focus dump instead of AS during haste trinket procs.
2) It increases your autoshot frequency (when not casting AIs)
3) It allows you to perform more SSs (when unhasted) and AIs during the CA phase when both have the high crit rate.
4) Increases pet attack speed

Yes, I understand the benefits in theory, but they aren't actually benefits when you can't use the focus you gain. You are almost certain to proc at least once per cycle, sometimes three. What do you do with the focus?

Sometimes you just end up spamming 3 Arcane Shots back to back just to get rid of focus, more haste means more focus from Steady Shots, from passive regen and even more procs.

Whytefyst,

The extra haste model that you are proposing at the end of your post above - would this be Paragon's 1850 haste setup as you discussed in post #511?

http://elitistjerks.com/f74/t112408-...1/#post1980851 ?

First, I want to be clear that I am not necessarily proposing an extra haste model for all players at all gear levels. It is an alternative approach for those that have sufficient gear to support it and want to try to gain some additional DPS from all of the focus generated by the T12 4-set.

The Paragon haste setup does add in additional haste over the 12.93% haste from gear and Pathing since it is 17.87% haste total from both the haste rating on gear and 3/3 Pathing. However, not including any haste trinkets, the haste is not sufficient to make any major changes to the rotation. It neither gains an extra GCD per cycle since SS cast time is only reduced by 0.06s from 1.4s to 1.34s nor does it lower AI cast time sufficiently to make it more worthwhile to cast since cast time is only reduced from 2.03s to 1.95s. This is shown in the table below:

	Unhasted	Non-H Trinket	H Trinket
SS cast	1.400	1.250	1.233
P SS cast	1.341	1.198	1.182
AI cast	2.030	1.813	1.788
P AI cast	1.945	1.737	1.713

Even when factoring in the haste trinkets, the difference between the 12.93% haste and 17.87% haste cases are small. The 12.93% haste case is exactly ideal for gaining an extra GCD per cycle with a non-heroic haste trinket proc since it lowers the SS cast time to 1.25s so you can now do 4 SSs in 5s. The 17.87% haste level just lowers the cast time by another 0.052s. In either case, the difference in AI cast time is about than 0.075s, hardly a significant amount to base decisions on whether to cast AI or not. Hence, I do not see much of a change in specials performed between the two cases.

Assuming a 2.9 speed ranged weapon, the difference in autoshot frequency is the same as AI cast time, so nothing really significant there either since it takes about 13.3 autoshots or about 24s to gain an additional autoshot. This does equal 12.5 more autoshots per 5 min fight; however, with casting a significant amount of AIs, the number of autoshots gained is somewhat less. I will estimate 8 more autoshots. That is more chances for WQ and T12 4-set procs but at slightly reduced damage.

Thus, as I stated in the previous post, I do not see much difference between the two set-ups and which you go with is preferential and dependent on your current gear ilevel.

The extra haste situations I refer to in the post a couple back are ones that add significant amounts of haste to either gain a GCD every CS cycle due to reducing SS cast time by 0.25s per each of 4 SSs or reducing AI cast time significantly.

The extra haste model I was referring to are the ones similar to those in the Max DPS thread. The case in that thread is the extreme case where the haste rating is greater than the crit rating ensuring that Matrix Restabilizer procs haste always instead of crit. With both haste trinkets procing at the start of the fight and using RFs, 1s AI cast times allow a lot of AIs to be cast in the CA phase.

Although this probably belongs in the Hunter Q & A thread, I'd like to post it here and get your take on it Whitefyst, seeing as you don't seem to be as involved in that thread as this one.

Simple question: If one is able to reach 12.93% haste from gear with 2/3 pathing, is it more beneficial to spec 2/3 pathing and 3/3 frenzy or spec 3/3 pathing and 2/3 frenzy and gain more mastery from gear instead of having to reach a higher haste rating from gear?

I must admit that I had not considered this option.

The result seems situational depending on your gear and pet WH uptime. For my character it was a 29 DPS loss. But then again I still have 2 359 pieces and no 391 pieces yet