Not sure if I fully understood the point of this, but is the purpose of this post to tell people to optimise around a buff with 17% uptime?
Washed up ex-Avast Raider. Still fucks around with Mythic+ | Signature by Pangcakes
Nah, the point of this is to show that haste actually gives more HoPo than expertise after soft cap and if you are using the SW talent the difference is even bigger.
---------- Post added 2012-11-30 at 04:44 PM ----------
In fact so big that even at my haste values and expertise values haste pulls ahead while using the SW talent.
Naturally the drawback is that you still get chances to get parried on CS .
---------- Post added 2012-11-30 at 04:46 PM ----------
and the purpose of my posts is never to tell people what to do. There are so many different ways to gear a protadin that saying 1 is correct is blatantly wrong. You can make educateded guesses what works best for the majoriy, but the fact is that so many factors scale in on how you should gear. Your current progress, your raid size, your other tank, your healers, your dps, your own playstyle etc etc. This was just to highlight a thing that is not very well known in the paladin tank community.
Different tanks are good at different things. Druids can keep up a high amount of dodge for a long time or Heal themselves fairly well. DKs can heal themselves extremely well and stack absorbs to preempt large hits. Pallies have a lot of CDs and can mitigate a tremendous amount of damage for a short period, by far more than any other class with cooldowns stacked.
Made sure of it by looking through several logs of a friend who has about one and a half percent expertise, and judging a boss dummy while naked for several minutes. Though I suppose I'd be interested in logs that say differently.
Last edited by Elathi; 2012-11-30 at 10:58 PM.
Except the point of hard capping exp is not to maximize HP generation. The point of hard capping exp is to prevent dmg spikes by ensuring that you will always have HP when you need it.
haste is winning in his analysis before you even take into account the other things haste effects ( like sacred sheild bubble re-procs and time till first bubble) and more seal of insight procs from more white swings as well as higher tps from spells like holy wrath which is also effected by haste.
also tested my previous theory with spell haste and seems it's back onto melle haste on the things i looked at ( cs/judge) joined some random dungys but never rolled a dps with the 10% melle speed buff to see ( even if only for myself) if it works on the cd's.
Why even soft cap expertise if you're going with this strategy? Just go for the 3% you need to cap spell hit.
We have no way to review the accuracy your claims about HoPo generation per point of stat, so the OP is essentially useless.
Basically, as many others and even myself stated in the OP, expertise is not useless, and it still gives reliability over throughput. Which is why expertise is probably better for those doing the bleeding edge content. However, for the majority of the wow community that is in early heroic modes or normal modes. A missed CS will not mean anything. Basically, anyone claiming missing a CS at normal modes and early heroic modes is the difference between life and death, is just blatantly wrong, as I have seen people doing full normal mode clears without having a clue about their rotation, basically having 50% of the normal HoPo regeneration, in which case a missed CS really is a piss in the ocean.
So yes, in some cases expertise is better for reliability over throughput, but this is not true in a majority of the cases, and the thing is, for those players where that 1 missed CS=death, well, those players should know this already and should not need that explained to them.
Expertise still remains the best stat after haste, and getting soft cap exp even when reforging and regemming everything to haste is very very easy.
As said earlier, it is very possible to find a balance where you think you are in good situation with throughput / reliablity.
Which is why I preferred to atleast soft cap as this put the throughput and reliability in a good spot for me.
Lets say you are at 10% expertise. Your chance to miss a CS is 5%. Your chance to miss 2 CS in a row is 0.25% Your chance to miss 3 in a row is 0.0125%
At 14% expertise, your chance to miss is 1%. Your chance to miss 2 in a row is 0.01% ( already lower than 3 in a row in the previous example ). 3 in a row would be 0.0001% chance.
So, at higher levels of expertise, it becomes extremely reliable even without capping. For 99.9% of the player base, missing one CS will not matter and for 95% missing 2 will not matter either. Though even at 10 expertise, the chance to miss 2 in a row is only 0.25%. Question is if it justifiable for most players to prioritise more expertise at this point over alot of throughput.
As others also mentioned, this is discounting all the other benefits of haste like faster SS procs, more SoI procs, using US, shorter LoH, DS, DP cooldowns.
---------- Post added 2012-12-03 at 05:32 PM ----------
Okay, so lets go into some of the math behind this and so many have asked.
First thing to do, is to establish how, do we calculate our HoPo regeneration?
A = B*(1-m)*(1+h)
Where B is our HoPo regeneration and M is our chance to miss and H is our haste % will give us our actualy holy power generation.
Now we need to expand B in order to cover the facts that CS is affected by expertise and that CS can proc GC, effectively giving each CS 1.2 HoPo.
Our HoPo regeneration could be stated as.
A = C*(1 - m)*(1+h)*1.2 + J*(1+h)
Where C and J are how much HoPo per second CS and Judgement generates.
Our crusader strike simply generates 1/4.5 HoPo at 0% haste. Our judgement however is rotationally limited to a 6.5s effective CD due to it clashing with CS every third CS, giving it 6.5s effective cd instead of the normal 6s. So our judgement regenerates 1/6.5 HoPo per second at 0% haste.
So C = 1/4.5
J = 1/6.5
Now to calculate our HoPo regen when stats changes.
Where dm and dh are added expertise or haste.
A = C*(1 - m + dm)*(1+h+dh)*1.2 + J*(1+h+dh)
Expanding this and ignoring second-order terms (dh*dm), we get:
A = 1.2*C*(1-m)*(1+h) + J*(1+h) + 1.2*C*(1+h)*dm + (1.2*C*(1-m)+J)*dh
The last two terms are the important thing here. Ae=1.2*C*(1+h)*dm and Ah=(1.2*C*(1-m)+J)*dh.
However since 1 expertise is 340 rating and 1 haste % is 425 rating we need to consider that.
Our goal is to learn how much each rating, r, gives, and since dm=r/340 and dh=r/425, we can write:
So now we can begin to play.
For h=0, m=0.075
Ae = (7.8431e-4)*r
Ah = (9.2897e-4)*r
Basically haste being 18% better for throughput at 0% haste and 7.5 expertise, and this scales quite linearly until 18.44% haste. Now this is of course only talking through put, remember that expertise still reliability, it is the everlasting throughput vs reliability discussion. Ofc discounting the other benefits of haste of expertise.
Now that covers the first part where you are not using SW.
Now to calcuate SW we need to find out the HoPo regen when using SW.
Ae = (nAe*(1-AW))*(awAe*AW)
Ah = (nAh*(1-AW))*(awAh*AW)
where n stand for normal rotation, aw for rotation during wings, and AW is the uptime for AW.
can we used for AW rotation aswell, except for the rotation changes during AW. During AW your rotation changes from CS-J-X CS-X-J CS-X-X
J-CS-J-X is the rotation during AW with SW, effectively putting CS on a 4 GCD CD.
So J = 1/3
C = 1/6
so we get
where J is 1/3 and C=1/6
Rest is the same as before.
---------- Post added 2012-12-03 at 05:59 PM ----------
To do a bit of a conclusion. The value in expertise increases the more damage a boss does. ( because the spikes will be higher ).
Basically anyone pushing heroic modes undergeared.
However, in these cases, stamina will always be better than expertise either way. I think that in any case that you can afford gemming secondary stats over stamina, you are geared enough to gem haste over expertise. But that is my opinion.
Fights where the damage is consistent favors expertise, however fights where you can anticipate the damage ( i.e. Sha of fear with Thrash ) or any fight where there is one phase where the damage is ramped up. ( i.e. WotE titan gas, Empress adds ) favors haste if you are using SW. If you are using HA that is another story but that is not the point of this thread.
Personally, I feel like most fights the damage is not so much RNG, you always know where the damage is coming and can prepare for that. Which in turn devalues expertise, since you can bank up holy power for when you actually need it. Most points of fights where tank dies is not during normal rotation.
Vizier - 3 stack exhales
Windlord - The tank damage attack whatever it is called
Garalon - Tank damage negliable, still cleaves.
Bladelord - Consistant damage, but still very low tank damage.
Amber-Shaper - Tank damage negliable
Empress - Bomb explosions, P2
So, in HoF, 3 bosses you can very much anticipate the tank damage and prepare for it. 1 boss the damage is very consistant, though IMO, not high enough still. 2 others where tank damage is not an issue, you could probably do those fights without ever using SotR.
MSV is another story since so many of the fights are magic damage bursty, so using exp or haste does not really matter.
Same goes a bit for terrace as first 3 bosses is almost only magical damage.
Sha damage is very predictable.
Though magic damage fights naturally favors haste over expertise.
Last edited by Firefly33; 2014-04-24 at 08:17 AM.