Remove Advertisements

[Split] Rotation Brainstorming/Discussion

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

[Split] Rotation Brainstorming/Discussion

Postby knaughty » Fri Mar 25, 2011 6:02 pm

Here's what I'm harping about re the protection rotation being annoying as hell - this is how a tank should be able to work out his rotation:

Our protagonist is a Ret paladin who just hit 85. He learns all his abilities, buys 2nd talent spec, decides to go Prot. He spends 5 minutes carefully building a prot spec by reading talent tool-tips, then another 5 reading ability tooltips and putting them on his bars, testing a few things on dummies as he goes. He's smart and not afraid to pull out a calculator to add up some numbers, but he's never heard of MATLAB, let alone having the skill to programatically simulate a tanking queue.

<Tries to use a variant of his Ret rotation as a tank> Oh WTF... Inquisition only lasts 12 seconds? And look how slow I get Holy Power! Do I use Inquisition every 3 HoPo or do I use ShoR?

<Thinks> Holy crap, ShoR hits like a truck, I guess that's my main threat move. My Ret big abilities were also on Holy Power, so that makes sense. Now, how to use it:

Q1: Do I do CS-ShoR-CS-ShoR or save for 3 HoPo?

OK, read tool-tip:
Code: Select all
Slam the target with your shield, causing Holy damage.  Consumes all charges of Holy Power to determine damage dealt:

1 Holy Power: (610 - 1) damage
2 Holy Power: (610 * 3 - 3) damage
3 Holy Power: (610 * 6 - 6) damage
OK, two HoPo is triple the damage of one, and three HoPo is double again. Looks like I should use 3 HoPo ones.
60 seconds testing on a target dummy
Right, 3 HoPo ShoRs are best.

<Thinks> I earn HoPo pretty much just from CS. There are hardly any procs like I had with Ret, so I have to spend a while getting 3 HoPo together... how do I do that?

Q2: OK, looks like CS-X-CS-X-CS-ShoR is my basic rotation, what's the best "X"?

OK, my choices are what... Judge, Avenger's Shield, Holy Wrath, Consecrate. What do they hit a dummy for...
30 seconds testing
Holy crap, AS hits like a truck, Judge is pretty good, Consecrate and HW are pretty shit and use a lot of mana. I guess they're for AE tanking only.
Reads talent tree and tooltips. Whacks on a dummy for a couple of minutes.
OK, I probably want to Judge on the pull to make sure the boss is slowed by JotJ, because that looks like -20% to my damage taken. Other than that, AS looks tons better - it can Proc Sacred Duty so that shield-bash is going to hit like a truck, and it sometimes has Holy Power so I can do the SotR sooner. So my rotation is:
  1. Judge to pull.
  2. CS-X-CS-X-ShoR
  3. AS > Judge for the "X"
  4. Hmm... sometimes I run out of this to cast, especially if a CS misses and I don't have 3 HoPo yet..

Q3: OK, what do I use for my dead GCDs?

Play with dummies for 5 minutes

Consecrate just makes me go OOM and has a really annoyingly long CD. I'll save it for AoE. So my dead GCDs are filled by Holy Wrath, so long as I have enough mana. And if I'm under about half mana, I really need to Judge a couple of times to get it back up.


Q4: Hmm... AS hits way harder than CS. If the proc is up, I assume that I use AS+>CS?

<Thinks, maybe pull out a piece of paper...>

(1) AS+ - CS - X - CS - ShoR on 5th GCD
(2) CS - AS+ - CS - ShoR on 4th GCD

Oh wow, that's interesting. Even if the proc is up, I don't skip my CS, I just wait for the next "X". Damn, I'm good for working that out!

Answer!

My ability queue is: ShoR > CS > AS > J > HW!
Last edited by knaughty on Fri Mar 25, 2011 6:07 pm, edited 1 time in total.
This isn't the "Offtankadin" forum. My MoP FAQ: http://tinyurl.com/FAQ-5-0
- Knaughty.
User avatar
knaughty
Maintankadonor
 
Posts: 3558
Joined: Mon Dec 17, 2007 10:06 pm
Location: Sydney, plotting my next diatribe against the forces of ignorance!

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby knaughty » Fri Mar 25, 2011 6:06 pm

Someone who works all that shit out and executes it properly should do significantly more damage than ArthåsPåladîn who hits random buttons to 3 HoPo then smashes ShoR because he like big numbers.

Instead we have 9 different queues that all do the same damage....
This isn't the "Offtankadin" forum. My MoP FAQ: http://tinyurl.com/FAQ-5-0
- Knaughty.
User avatar
knaughty
Maintankadonor
 
Posts: 3558
Joined: Mon Dec 17, 2007 10:06 pm
Location: Sydney, plotting my next diatribe against the forces of ignorance!

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Sur-Pseudo » Sat Mar 26, 2011 2:57 am

Yeah, I didn't want to say anything but..

We have sims for 30 different situations, which basically boil down to like.. 8.... and that's more around finishers and would be better explained then listed (e.g. "If you have high Hit/expertise, you will find Inq -> SotR works slightly better then the SotR>CS>AS>J>Filler listed")..

Fillers seem to basically not matter.. and what's left is a block of 30 sims that are cryptic to anyone that doesn't follow this thread on a normal basis (hell, I disappeared for a week and AS+/AS++/Inq*/sdSotR came outta nowhere, and sd)


I'd say reducing the listed priority lists would help on multiple fronts:
1) Reduced time to generate sims
2) Easier to read & Understand, less need to create complex shorthand to try and fit a table... can explain a few specific sims (as opposed to explaining many more out of 30 sims)
3) Shows people they don't need to find their specific priority que, and that in many cases priority does not matter (after a certain base layout)

Perhaps depreciate some sims and list them at the end as "These sims proved to be statistically close to similar sims above"... or "If the sim is not listed and is similar to the above, it's been shown to be insignificantly different for dps"
Sur-Pseudo
 
Posts: 117
Joined: Sat Apr 05, 2008 9:44 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby knaughty » Sat Mar 26, 2011 3:25 am

Sur-Pseudo wrote:Yeah, I didn't want to say anything but..

We have sims for 30 different situations, which basically boil down to like.. 8.... and that's more around finishers and would be better explained then listed (e.g. "If you have high Hit/expertise, you will find Inq -> SotR works slightly better then the SotR>CS>AS>J>Filler listed")..


Ah, this I disagree with.

Theck has do do his eleventy-seven simulations so that we can see what the DPS of each rotation is - yes, our top 9 are all the same. But you don't know they're going to be the same until you do the sims.

The problem is the tuning of the abilities, not the number of simulations that Theck considers.
This isn't the "Offtankadin" forum. My MoP FAQ: http://tinyurl.com/FAQ-5-0
- Knaughty.
User avatar
knaughty
Maintankadonor
 
Posts: 3558
Joined: Mon Dec 17, 2007 10:06 pm
Location: Sydney, plotting my next diatribe against the forces of ignorance!

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Sur-Pseudo » Sat Mar 26, 2011 3:42 am

true, if the decide to adjust abilities we need to sim them, but we should also remember not everyone trolls this thread, and will be better served with an easier layout

If the time to sim is okay, sure, keep going, but we need to explain that sdAS/AS++++ means whatever, or we're asking someone who just joins the thread to find the page number if references (For example, I didn't know what AS++ meant as I was gone for a week, luckily It was only 5 pages back, as the forums don't let you search for "AS" (and the plus' are ignored by search))



and Knaughty, your signature hurts me... I am an offtankadin :-(
Sur-Pseudo
 
Posts: 117
Joined: Sat Apr 05, 2008 9:44 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby knaughty » Sat Mar 26, 2011 4:09 am

Sur-Pseudo wrote:true, if the decide to adjust abilities we need to sim them, but we should also remember not everyone trolls this thread, and will be better served with an easier layout

<snip>

and Knaughty, your signature hurts me... I am an offtankadin :-(


(1) I'll write up a simple version of our single-target and AE rotations for people who can't cope with thte pounding headache that this thread represents.

(2) That bit of my sig dates back to TBC, when tankadins as a class where entirely dismissed as off-tanks. I took real pride in going "Screw that, I'm going to goddamn main-tank Tier 5, 6 & Sunwell!" This was a seriously unusual thing at the time. If you're an off-tank that happens to be a paladin, it's a much smaller deal nowadays because paladins are actually just one of the four tanking classes.
This isn't the "Offtankadin" forum. My MoP FAQ: http://tinyurl.com/FAQ-5-0
- Knaughty.
User avatar
knaughty
Maintankadonor
 
Posts: 3558
Joined: Mon Dec 17, 2007 10:06 pm
Location: Sydney, plotting my next diatribe against the forces of ignorance!

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Sat Mar 26, 2011 11:06 am

So, some general comments re: the change.

I don't disagree with you about how disappointing it is that there's no significant difference between any queues. However, that has very little to do with this modification to Sacred Duty, and it was true before this change anyway. It's root is the 3-second Crusader Strike. We're partly to blame for that, because we complained rather loudly about the 4.5-second version, but I've been playing with some ideas lately and to be honest, there are very few ways to make any significant skill-based variation within the confines of a 3-second heartbeat.

If you really want to have a dynamic, interesting rotation, we need to go back to the 4.5-second Crusader Strike model. No matter how you tweak things, the HP resource mechanic ends up either dominating or being completely irrelevant. The latter case is obviously not attractive for design reasons, so we inherently want to be in the former situation.

If the resource mechanic dominates, then CS>all is all but guaranteed. With a 3-second cooldown, you only have one GCD to play with, which severely limits what you can do to make the rotation dynamic. It will only ever become a question of "which filler is better," because you're not capable of coming up with situations where the order of that filler matters.

It's worth noting that Ret is also a slave to the CS-TV cycle, but it plays really well. It doesn't feel confining at all, and I think the major reason is that there's a 2-GCD gap between CS's. That extra space means you can do more interesting things with procs and more seamlessly weave other abilities in and out. And the button presses do matter in the Ret rotation. The week after 4.0.6, I was out-dpsing our main-spec Rets by a good 20%, despite still having a few blues in my setup. Why? Because I was properly following the new queue that was optimized for the new HoL/DP mechanics, while they were still following the 4.0.3 queue. It was very apparent, and the next week (after I pointed it out to them and they had time to practice it), they were back to being neck-and-neck with me or beating me.

With 4.5 seconds between CS's, you can do a lot of interesting things. Not just in terms of procs to respond to, but in terms of having a slightly more complicated set of priorities. The fact that you can cast Exo and Judgement in any order matters, because you can tweak things to make that order matters. In short, we just don't have the free GCDs in the 3-second model to work with, which is the root problem.

Note that the Sacred Duty change probably isn't intended to fix this. I think it's fairly clear what they've been trying to do based on the progression of changes. They're trying to make Grand Crusader a decent talent. You've said it yourself in the basic guide - Grand Crusader in 4.0.6 and earlier is just meh for a number of reasons. But the primary one is that AS is in the last spot of our priority queue (constraining ourselves to the four "core" abilities right now).

Why is it in the last spot? Sacred Duty. Ever since 4.0.1, we've seen that J>AS is dominant, and it has always been because of Sacred Duty propping up the effective damage per cast of Judgement. So, I'm going to go out on a limb here and make some guesses based on the changes we've seen:

  • Problem: Grand Crusader bad, to the point that some people are skipping the talent.
  • Solution #1: Maybe the rotation is too tight. Nerf Holy Power on CS, which opens up a few more free GCDs. Now we have GCDs to take advantage of those Grand Crusader procs.
  • Result: 4.0.6 nerf to HP generation, my sims still put AS in last place. Oops. Despite having more GCDs, we still don't prioritize AS, and sit on the proc.

  • Problem: OK, that didn't work. And now they're complaining that HP generation is too slow. What if we put HP generation on AS somehow? (Note that this is one of many things that you and I suggested as far back as beta in various threads.)
  • Solution #2: 4.1 version of Grand Crusader. Holy Power generation on AS definitely makes the GC talent more attractive.
  • Result: Sims still show J>AS better. So despite being more attractive, we're still going to sit on procs and only make use of them when there's nothing better to cast. Since this information could only have come from this thread, we can feel pretty confident they're watching us (spooky!). Which also means that they have a clue about the reason, since I've said on several occasions now that SD is really the thing keeping J ahead.

  • Solution #3: Well, if Sacred Duty is the only thing keeping J ahead.... what if AS also procs SD? Now there's no advantage to AS over J as far as that's concerned, which does a number of things.
  • Results:
    1. AS>J now reasonable
    2. Much less likely to sit on Grand Crusader procs
    3. From the sims, it seems there are even cases where AS+>CS is at least DPS neutral, if not a slight gain. I bet if we dig into this more deeply, we can come up with some really neat situational queues that show a slight increase, since this situation only occurs around SotR casts (i.e. in general you get CS*-AS and CS>AS vs. AS+>CS is irrelevant, the only time we care is if it's CS*-SotR-??).
    4. Downside: it brings a bunch of queues more or less in line with one another in terms of DPS. SotR>CS>(whatever) works. This was pretty much the case before though, J>AS was only ever about 100 DPS ahead of anything else, which is small enough that it's irrelevant for most players.

Again though, there's very little you can do to make filler spells have a huge impact if you're shackled with a 3-second resource generation heartbeat. I've been trying to think of some ideas that fix this, but they all involve either neutering CS/SotR to a ridiculous degree or completely reworking Sacred Duty in one of a variety of ways.

Examples:
Joverpower - Sacred Duty resets cooldown on Judgement and/or gives it a damage buff when your CS is dodged.
Jrevenge - Sacred Duty resets cooldown on J and/or gives it a damage buff when you dodge/parry
SDeclipse - Sacred Duty gives an eclipse-like effect, that buffs AS if your last filler was J, and buffs J if your last filler was AS. Might be interesting, rewards players who are attentive and choose the filler appropriately.
SDhpg - SD now triggers off of CS/HotR, grants auto-crit to Judgement and possibly resets J cooldown.

You could also imagine adding a 5th spell to the mix, ala Art of War. Any of the above examples would work with instant-cast Exorcism as the spell being "awarded." But again, that works best in a situation where you have more than one GCD between holy power generators - with just one GCD to work with, it's very, very hard to come up with something dynamic.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7759
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Yelena » Sat Mar 26, 2011 3:59 pm

theckhd wrote:It's root is the 3-second Crusader Strike. We're partly to blame for that, because we complained rather loudly about the 4.5-second version, but I've been playing with some ideas lately and to be honest, there are very few ways to make any significant skill-based variation within the confines of a 3-second heartbeat.

If you really want to have a dynamic, interesting rotation, we need to go back to the 4.5-second Crusader Strike model. No matter how you tweak things, the HP resource mechanic ends up either dominating or being completely irrelevant. The latter case is obviously not attractive for design reasons, so we inherently want to be in the former situation.

Part of the problem stems from the fact that everything in our "rotation" has a cooldown. Being that our primary resource generating mechanic also has a cooldown, it ends up being the overall heartbeat of our rotation, which by association is going to limit the options available by how much "free time" we have between beats. This is one of the reasons several of us were pushing for some form of spammable ability, as that does open the avenue to more dynamic rotation options. As it stands now, the skill floor required to play a Prot Paladin is rather low, as one has to legitimately try to screw the rotation up; nevermind the fact that there have been plenty of examples of that happening, but the same could be said for all four tank classes.

I agree that we are partly to blame for the current state of our rotation, due to our rather negative feedback during beta regarding the 4.5 second beat. The fact that it doesn't really matter what we use for fillers also ties into the constraints a three second beat leaves us with. With the kind of rotation model we have, from a development standpoint, your options basically boil down to:
  • Making one ability significantly better than everything else, to the point one would be foolish to use anything else, unless that significantly better ability is on cooldown. (Pre-4.1 J, due to SD procs)
  • Making your "options" balance out, such that one really doesn't have a significant advantage over another.
One could argue in favor of either or neither, but when you're effectively limited to using one filler between beats, those options are what it boils down to. I don't necessarily agree with this, as I was one of the proponents toward the "spammable filler" style of rotation, but this is what we got, so I dealt with it.

----------------

I'm also somewhat concerned with the current implementation of our Mastery. Unless the proposed "raid boss expertise" becomes a reality, and scales decently with each tier, the relative value of Mastery is going to drop off sharply as item level increases; as our effective hardcap is already reachable in current content if one tries to. The obvious counter-approach to this is to reforge it away, once the effective hardcap is reached, which potentially exacerbates the reverse-scaling issue we've already started to see with Vengeance (the current implementation of Vengeance being another issue altogether).

None of this is new, of course, we were bringing this issue up during the beta as well.
Valleri - Frostwolf
Yelena
 
Posts: 335
Joined: Sat Jun 27, 2009 4:54 am
Location: Behind the Eight Ball

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby tlitp » Sun Mar 27, 2011 3:34 am

Yelena wrote:
theckhd wrote:(...)how disappointing it is that there's no significant difference between any queues.(...) It's root is the 3-second Crusader Strike.
(...)
If you really want to have a dynamic, interesting rotation, we need to go back to the 4.5-second Crusader Strike model. No matter how you tweak things, the HP resource mechanic ends up either dominating or being completely irrelevant.

Part of the problem stems from the fact that everything in our "rotation" has a cooldown.

  • 3-sec HPGs are OK, but not in the context of the 4.0.6 behaviour; killing off any sort of Inq-weaving is not acceptable as it makes the cycles much more rigid.
  • 4.5-sec HPGs are also OK, as long as the finishers aren't excessively prioritized. The proposed 4.1 implementation of SD is not OK, as it throws SotR off balance.
  • the HoPow-scaling is not OK; as already stated six months ago, "using a 3-HoPow finisher should be a decision - optimal or less optimal - not something that is deemed mandatory in any given circumstance". Inq needs to be at least 6/9/12, if not 8/10/12; SotR needs to drop the 1/3/6 absolute scaling (i.e. on both base and AP coefficient) in favor of a standard 2/4/6, but only on the AP coefficient.
  • at this point in time, we can safely state that there is absolutely no reason to have a cooldown on HW : relatively high cost, poor scaling (only off SP), inverse AoE scaling. Let's be serious, dropping the cooldown won't make the Holy Paladins killing machines... but it'll turn HW into a better filler for both Prot and Ret. Evidently, the corresponding glyph has to be changed as well; reducing the mana cost seems to be the better alternative.
  • again (and again, and again, and again...), the ability toolbox is, currently, ridiculously designed; buffs/debuffs/support (JotJ/Vind/JotW/interrupt/silence) should not be linked to standard abilities (J/AS). Rogues don't stun/silence/debuff via Crusader Strike, for fuck's sake. :mrgreen:
  • lastly, the talent matrix is not OK; having abominations such as WotL (or the proposed 4.1 SD) that simply scream "pick me" is anything but good design. Refer to the rest of the damage-oriented options; they're much more balanced, allowing the Paladin to juggle them at will.


@Theck : interesting discussion, but somewhat misplaced. Do you mind creating a distinct thread for it, for improved visibility and easier referencing ?
User avatar
tlitp
 
Posts: 556
Joined: Mon Jul 27, 2009 3:25 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Mon Mar 28, 2011 8:06 am

tlitp wrote:@Theck : interesting discussion, but somewhat misplaced. Do you mind creating a distinct thread for it, for improved visibility and easier referencing ?

Done. This thread can be for discussing the pros/cons of any detail of the rotation, as well as brainstorming about ways to "fix" or improve it.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7759
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Theck's MATLAB thread - Cataclysm/4.x

Postby Cavalorn » Mon Mar 28, 2011 8:13 am

First of all, reading and understanding is, for me, much easier than writing in english ;)

Just from my point of view:
If you get headache from reading this threat and/or don't like to compare dps-cycles which are "roughly" the same. Go ahead and read the Tankadin-Basic-FAQ. Everything, which is important for you, will be explained well enough there....but you are right, if you want to understand all the posts you have to read nearly every post. No chance to skip a few pages.

On the other side, if you like crunching numbers(with or without knowledge of matlab), if you try to understand why cycle xy will - under the given circumstances - perform better than the other ones, yes, here you are! In this thread you will find the infos you are looking for :D

I think der is no reason why theck & tlitp should cut the output of there hard work down to a "easy to understand version" and waste tons auf simulations. Basic-FAQ and the "TLDR" should be enough for this.

Edit: hmm a few posts are gone, perhaps my reply is now at the wrong place ;)
Cavalorn
 
Posts: 18
Joined: Tue Apr 14, 2009 4:00 am
Location: Austria

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Mon Mar 28, 2011 8:18 am

tlitp wrote:
  • 3-sec HPGs are OK, but not in the context of the 4.0.6 behaviour; killing off any sort of Inq-weaving is not acceptable as it makes the cycles much more rigid.
  • 4.5-sec HPGs are also OK, as long as the finishers aren't excessively prioritized. The proposed 4.1 implementation of SD is not OK, as it throws SotR off balance.
  • the HoPow-scaling is not OK; as already stated six months ago, "using a 3-HoPow finisher should be a decision - optimal or less optimal - not something that is deemed mandatory in any given circumstance". Inq needs to be at least 6/9/12, if not 8/10/12; SotR needs to drop the 1/3/6 absolute scaling (i.e. on both base and AP coefficient) in favor of a standard 2/4/6, but only on the AP coefficient.


To be fair, I think they've made it pretty clear that they don't want Inq-weaving to be part of the single-target rotation. So killing that off may have been part of the goal. One of the repeating themes we've seen in blue posts is that they don't want the rotation to require too much attention. We're on the other end of things, in that we're asking for a much more complicated rotation to make it interesting. But in making that request, we might not be speaking for the majority of tankadins.

With a 4.5-sec HPG cycle, finishers are automatically not as highly prioritized, because you have 2 GCDs in which to use them. That said, SotR misses would probably make SotR prioritization "better" isnofar as it would reduce the penalty for whiffing. The 4.1 SD implementation probably works even better in the 4.5-second cycle, because things could be tuned to make SDSotR>sdAS>sdJ>AS>SotR>J, for example. Keep in mind that with the move to a 4.5-second HPG cycle, damages/cooldowns/procrates would all have to be tweaked. The point is that with 2 GCDs available you have choices as to what you cast and in what order.

How does the scaling on rogue finishers work? Is it completely linear with CP?
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7759
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: [Split] Rotation Brainstorming/Discussion

Postby bldavis » Mon Mar 28, 2011 8:41 am

i generally avoid posting in theorycraft threads, but theck asked about rogue finishers and cp scaling


Slice and Dice
S&D is a linear buff, more cp = longer S&D

Eviscerate
evis also appears to be linear scaling per cp

Rupture
Rupture appears to be non-linear (prefered finisher for some specs)
however this is just looking at the tooltip, i dont have the time to do the math right now

these are the main three you would see a rogue use
they could add others for raid debuffs, or depending on spec, but these tend to be universal
Image

Brekkie:Tanks are like shitty DPS. And healers are like REALLY distracted DPS
Amirya:Why yes, your penis is longer than his because you hit 30k dps in the first 10 seconds. But guess what? That raid boss has a dick bigger than your ego.
Flex:I don't make mistakes. I execute carefully planned strategic group wipes.
Levie:(in /g) It's weird, I have a collar and I dont know where I got it from, Worgen are kinky!
Levie:Drunk Lev goes and does what he pleases just to annoy sober Lev.
Sagara:You see, you need to *spread* the bun before you insert the hot dog.
User avatar
bldavis
 
Posts: 7350
Joined: Mon Sep 07, 2009 12:04 pm
Location: Searching for myself. If i get back before I return, please have me stop and wait for myself.

Re: [Split] Rotation Brainstorming/Discussion

Postby Chicken » Mon Mar 28, 2011 9:01 am

Technically Slice and Dice isn't entirely linear. The first combo point being worth 9 seconds, compared to any extra combo points adding 3 seconds (Barring the glyph, which changes it to 15/3, the relative value isn't modified by the duration increase talent as it keeps the same difference). Anyway, their linear scaling is countered in design through rogues being resource limited, and all their finishers consume both their primary resource (energy) and secondary resource (combo points), so anything used with a higher amount of combo points is more primary resource efficient. Which encourages rogues to usually use their abilities with a high amount of combo points.

Comparatively Paladin 'finishers' only consume the secondary resource (holy power), thus meaning that if the direct damage dealing ones were modified to scale linearly, the biggest reason to save up to three Holy Power would be to save GCDs. Which actually works out reasonably well since that's actually the primary limitation for Paladin ability usage anyway; the reason specific priority queues are the best are because they do the best job at filling as many empty GCDs as possible with the highest damage dealing abilities. It'd work particularly well for clearing up gaps in the rotation with a longer cooldown on Crusader Strike. I'd also expect it to lend a bit more depth to Sacred Duty usage, you'd always want to aim for SD being used on a 3-point ShoR when it does proc.

I expect Blizzard wants to keep the feeling of 3 Holy Power being what you build up to for a big burst however, which you'd lose with linear scaling.
Image
User avatar
Chicken
 
Posts: 1597
Joined: Fri Jun 26, 2009 2:19 pm

Re: [Split] Rotation Brainstorming/Discussion

Postby Petrus » Mon Mar 28, 2011 9:46 am

Going off of what Theck wrote about Ret feeling dynamic and good and this statement of his about rotation spells:

You could also imagine adding a 5th spell to the mix, ala Art of War. Any of the above examples would work with instant-cast Exorcism as the spell being "awarded." But again, that works best in a situation where you have more than one GCD between holy power generators - with just one GCD to work with, it's very, very hard to come up with something dynamic.


We've essentially got a 4-spell rotation (CS/SotR/J/AS) and Ret also has a 4-spell rotation (CS/TV/J/Exo) not counting Inquisition. The thing that makes ret smooth with a 4.5sec CS is that it prioritizes procs just enough that it can usually fill both open GCDs between CS hits (though on occasion one or both do end up open, it isn't common). I wouldn't mind a move back to a 4.5 CS but they'd have to change a proc somewhere - probably grand crusader again - to work off of autoattack like Art of War and not just off of CS. Either way it isn't a perfect solution and perhaps the proc to duplicate isn't necessarily Art of War but Divine Purpose. I suppose it would feel more Sword-and-Board-ish and almost too Ret-like, but having a 3HP SotR occasionally might be a nice way to complement a 4.5 second cooldown CS rotation.

I like the changes they're making because they clearly understand there's a problem, and that's a good thing. I'm just not exactly sure how good/bad they will be in the long run without making the rotation much more dynamic a la ret or prot warrior rotations.
User avatar
Petrus
Maintankadonor
 
Posts: 940
Joined: Wed Feb 06, 2008 3:45 am
Location: St. Paul, Minnesota

Next

Return to Advanced Theorycraft and Calculations

Who is online

Users browsing this forum: No registered users and 1 guest


Remove Advertisements

Who is online

In total there is 1 user online :: 0 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 380 on Tue Oct 14, 2008 6:28 pm

Users browsing this forum: No registered users and 1 guest
?php } else { ?