[Prot] 5.0 - I'm Sexy and I Know it - Page 6

Muhrat · 10/25/12, 10:50 AM

I see your point. Thanks! Do you recommend gemming towards haste? Using haste-stamina gems in blue or yellow sockets?

Wrathblood · 10/25/12, 4:41 PM

Originally Posted by Theck

I don't agree with your "mastery>haste" conclusion. First of all, haste will be flat-out better for DPS. So yes, you can do 200k+ DPS on heroic Stone Guard with a mastery build, but you'll still do more DPS with the haste build.

Second, I think your conclusion about mastery being better against magical damage is flat-out wrong. Yes, mastery buffs our Bastion of Glory heal. Which we can only use effectively if we're regularly using SotR, which we would more or less not be using on a boss with a high amount of magical damage. So, in fact, on a boss that does primarily magic damage, mastery is sort of useless.

Haste, on the other hand, gives you more HP generation and thus more WoGs (and if you do happen to be alternating, more BoG buffed WoGs). It also gives you more Seal of Insight procs (so more self-healing) and more Battle Healer healing on the raid. So it has a nonzero advantage on any type of fight where you use WoG, which makes it strictly superior to mastery.

Third, I think that Feng is a very poor example of a boss with high magical damage. Most of his magical damage is not dangerous to the tank - it's either a slow DoT or relatively weak compared to our health, and much of it is negated by the fight mechanics (Nullification Barrier or interrupts). What does kill you on Feng is his melee, specifically the double melee attacks during the shield phase. He will hit you twice for around 200k damage within 0.1 seconds fairly consistently through that phase, and in my experience this is the only part of that fight that actually endangers the tank. So I would argue that Feng is the textbook example of a boss with hard-hitting physical damage.

Being able to have SotR up for every other "double attack" makes a huge difference in how much spike damage you take, and subsequently in your healer's panic level and mana expenditure. Feng is one of the best examples of a boss where timing SotR casts carefully makes a huge difference, because a good tank can cycle cooldowns and SotR coverage to make sure they never take an unmitigated double attack. The haste build gives you more access to SotR, making it far easier to coordinate that coverage (SotR, SotR, cooldown for 8-10s, repeat).

Hmm, I actually disagree somewhat. I think Feng is an excellent example of a fight where Mastery shines. Not because its a magic-heavy fight (because, from a tank standpoint, I agree that its not, and beyond that I also agree that Haste is better for a magic-heavy fight anyway) but because its a fight with frequent and regular tank swaps.

Lets say that a paladin tank is tanking against Feng and the tanks swap exactly every 30 seconds (with Barriers and other mechanics this certainly isn't true, but its just for the sake of argument). When the paladin isn't tanking, that person is juggling their rotation to ensure that they have 5 HoPo and both CS and J off cooldown when their turn comes around again (essentially starting every 30 second phase at the start of their rotation).

By my math, you could probably land roughly ~13 HoPo generators in that time (including 1 GC-AS), plus the 5 starting HoPo, gives you ~6 ShoR's and 18 seconds of ShoR up-time. So 60% ShoR coverage while you're tanking Feng. Lets say you've got 5k itemization you can put into either Haste or Mastery, and buffed, you're getting 45% ShoR damage reduction.

The 5k itemization will give you either ~12% more Haste or ~8% more damage reduction on your ShoR buff (plus block and extra WoG healing).

Lets say you go with Haste and that additional 12% gives you 12% more HoPo and thus, on average over the course of the fight, ShoR buff-time. That's about 2 more seconds of ShoR time, moving our uptime from 60% to 67%. Multiply by the 45% damage reduction from ShoR, we've gone from 73% damage to 70% damage, a 4.1% reduction in damage taken.

Alternatively, lets say you go with the Mastery. 8% more damage reduction increases your damage reduction from 45% to 53%. With 60% ShoR uptime, you've gone from taking 73% damage to 68%, a 6.8% reduction in damage taken. In addition, you've increased your chance to block incoming attacks. I believe block suffers from DR so you'd probably increase your chance of blocking from ~27% to 34%.

Including the effect of block on the choices above, Haste goes from 70% (times 27% chance of block to reduce damage by 30%) to 64% damage taken. Mastery goes from 68% (times 34% chance of block to reduce damage by 30%) to 61%. In case this isn't clear, including Block increases Mastery's damage reduction advantage over Haste by about another 30%.

The base case then goes from 73% to 67%. So, switching to Haste reduces your damage taken by ~5% while putting those same points into Mastery reduces your damage taken by 9%. Mastery is ~80% better at damage reduction than Haste in this situation. Haste would reduce further the periods during which ShoR isn't up, but Mastery makes the ShoR buff so much stronger, plus increasing your chance of blocking, that it wins pretty easily.

That doesn't make Mastery better overall necessarily, but its certainly better at keeping the tank alive.

mrbreck · 10/25/12, 6:56 PM

In the case where we have a choice between a piece of "DPS" gear that has no crit and a "Tanking" piece that has reduced STR in favor of an increased amount of avoidance plus another tanking stat (hit, exp, mastery, or haste), which is the better option? The Klaxxi necklaces and pants are a perfect example of what I'm talking about.

apoptygma · 10/26/12, 1:26 AM

Originally Posted by mrbreck

The Klaxxi necklaces and pants are a perfect example...

Comparing Bloodseeker's Solitaire and Paragon's Pale Pendant -

I'd take the Paragon's Pale Pendant over Bloodseeker's Solitaire based on the fact that the smaller stat on Paragon's is hit which you are unlikely to want to reforge and the larger secondary stat is 638 Parry, meaning you can turn 255 of that into Expertise leaving only 383 as the 'undesirable' parry. Of course all these would be mute if you were already hit/exp capped elsewhere then Bloodseeker's is the better option.

My concern with this whole paradigm is that all the gear we are offered with the new looting system isn't itemized toward active mitigation which will end up with one of a few outcomes:

1) Most players continue to stack avoidance etc. and only the elite minority who read forums gear appropriately
2) Blizzard nerf Haste / Mastery into the ground and perhaps add some expertise and/or hit requirement reduction from talents/spells
3) Avoidance gets a massive buff to make it preferable compared to haste and mastery
4) Blizzard trash the whole concept and all our gear becomes worthless/off-spec

roughly in order of most to least likely...

Charybdis · 10/26/12, 2:33 AM

Originally Posted by apoptygma

My concern with this whole paradigm is that all the gear we are offered with the new looting system isn't itemized toward active mitigation which will end up with one of a few outcomes:

I'm not entirely sure if they intended the active mitigation stats to mathematically be better than dodge or parry, but thanks to how they've made such a big deal out of active mitigation I think it's very safe to say we'll continue to get gear choices with expertise, hit, and mastery. Given the diversity of plate DPS we'll likely see plenty of acceptable haste pieces too.

Haste will certainly get nerfed first if they feel it's too good for us since the stat isn't traditionally meant for plate tanks to get much from.

Onodrim · 10/26/12, 3:44 AM

Originally Posted by Theck

If you mean the effect that weapon procs will have on the debate between haste and mastery in general, then it's almost certainly insignificant. They will likely all scale roughly the same with haste, so it will still come down to the individual effects of each weapon enchant.

Indeed. As I understand the new 'Real PPM' system, your proc chance is calculated the following way:

PPM*Haste*Time since last proc chance/60

(with time since last chance capped at 10 seconds).

As with the previous 'PPM' system, Haste will increase the frequency with which your weapon enchants will proc. With this in mind, it could become relevant to pose the hypothesis: Potential added uptime of your weapon enchants favours Haste over Mastery as a general recommendation.

Originally Posted by Theck

I do plan on doing enchant modeling eventually just to give us some hard numbers on what exactly each enchant grants. It's not particularly hard to code, it just takes some time. We already have a rough understanding of the other enchant proc mechanics; last time we checked CS/HotR/J/AS/Melee were the only proc triggers for dynamic effects. That said, this was fairly early in beta, so it's possible things have changed; a re-test on the proc triggers would probably be a good idea.

Right. I'll see if I can find the time to do a re-test; if nothing else, I'd suspect Windsong procs from WoG.

Theck · 10/26/12, 10:59 AM

Originally Posted by Muhrat

I see your point. Thanks! Do you recommend gemming towards haste? Using haste-stamina gems in blue or yellow sockets?

I tend to use hit/stam in yellow sockets, in order to reforge more into haste. But haste/stam in a yellow socket is also reasonable, especially if the numbers work out such that it lets you reforge to hit-/exp-caps with less wasted overage. Blue slots get stamina, period.

Originally Posted by Wrathblood

That doesn't make Mastery better overall necessarily, but its certainly better at keeping the tank alive.

I disagree. You've just suggested that mastery>haste based on a TDR argument. Which is immediately invalid, in my mind, since neither of the control/mastery or control/haste gear paradigms is based around the concept of minimizing total damage. If I had a yellow flag to throw, I'd be tossing it at my monitor right now. :P Using that logic, you should gear for dodge and parry, because both of those are better than either haste or mastery for TDR!

Mastery will certainly buff your SotR, and may give you better TDR and a little block. But that 5 saved-up holy power isn't enough to make that matter; you're still going to get more SotR coverage, and more importantly have more control over that coverage, by using haste.

Especially given that the only real "tank-death" threat in that fight is the double-melee. You live or die based on how many of those double attacks you can cover with mitigation, either a cooldown or SotR or whatever. Ideally you cover all of them. It doesn't matter whether you mitigate 50% or 58% of those double attacks - in fact, Divine Protection is often enough with only 20% - it just matters that it's not 0%. So in that scenario, I would definitely take 2 more seconds of coverage via haste than 8% damage reduction via mastery - the 8% damage reduction doesn't turn the death condition into a live condition, but the extra coverage can.

(Aside: this is also the reason that I disagree with you about our tier bonuses. The 4-piece is good, in my mind, but the 2-piece is amazing. AD is a phenomenal "oh shit" button, and having it available more often is great. Whereas 10% extra mitigation on SotR is nice, but it's just more TDR reduction that probably doesn't much matter in any practical sense.)

Originally Posted by Onodrim

Right. I'll see if I can find the time to do a re-test; if nothing else, I'd suspect Windsong procs from WoG.

I'd imagine Windsong procs from everything now, which means that 2 procs per minute is a reasonable enough estimate (Hamlet has a post about the specific combinatorics you need to do to make that more accurate, so I won't reiterate them here). In any event, it's not Windsong that we need to test, it's Dancing Steel. We know how traditional PPM mechanics work, but we need to be certain about the proc triggers to accurately model its uptime.

Wrathblood · 10/26/12, 7:57 PM

Originally Posted by Theck

I disagree. You've just suggested that mastery>haste based on a TDR argument. Which is immediately invalid, in my mind, since neither of the control/mastery or control/haste gear paradigms is based around the concept of minimizing total damage. If I had a yellow flag to throw, I'd be tossing it at my monitor right now. :P Using that logic, you should gear for dodge and parry, because both of those are better than either haste or mastery for TDR!

Mastery will certainly buff your SotR, and may give you better TDR and a little block. But that 5 saved-up holy power isn't enough to make that matter; you're still going to get more SotR coverage, and more importantly have more control over that coverage, by using haste.

Especially given that the only real "tank-death" threat in that fight is the double-melee. You live or die based on how many of those double attacks you can cover with mitigation, either a cooldown or SotR or whatever. Ideally you cover all of them. It doesn't matter whether you mitigate 50% or 58% of those double attacks - in fact, Divine Protection is often enough with only 20% - it just matters that it's not 0%. So in that scenario, I would definitely take 2 more seconds of coverage via haste than 8% damage reduction via mastery - the 8% damage reduction doesn't turn the death condition into a live condition, but the extra coverage can.

(Aside: this is also the reason that I disagree with you about our tier bonuses. The 4-piece is good, in my mind, but the 2-piece is amazing. AD is a phenomenal "oh shit" button, and having it available more often is great. Whereas 10% extra mitigation on SotR is nice, but it's just more TDR reduction that probably doesn't much matter in any practical sense.)

We see it really differently. Our 2 piece is, if used perfectly, a 2.7% increase in uptime for 20% damage reduction (plus, of course, the no-death thing), while our 4 piece is 10% physical damage reduction that's up 40% of the time. Really its not about damage smoothing anymore, is it? Its about EH as your advice above kinda suggests.

This could be a difference in perspectives. In 10-man heroics, I can be spiked down, but my healers can also be overwhelmed. To me, simple damage reduction is less valuable than spike reduction, but it still has value. The way you're describing your stat prioritization, it honestly doesn't sound like Haste OR Mastery is the correct answer. If DR has no value and you're at risk of being burst down every ~3 GCDs by Feng, the correct answer is STAM because you're in Heroic ICC tank spike damage-land.

Edit - In fact, I wonder if there's an argument to be made for stacking STAM over, say, Exp. After all, if you're in 2-swing territory, capping Hit/Exp and stacking Mastery only cuts your chance of a 90% damage spike by about 1/3 over stacking Avoidance. In 3-swing territory, it cuts it about in half. Dumping 7600 itemization points into STAM instead of Hit/Exp would buy you a bunch of EH and I wonder if that might actually be more valuable.

Theck · 10/26/12, 9:24 PM

Originally Posted by Wrathblood

We see it really differently. Our 2 piece is, if used perfectly, a 2.7% increase in uptime for 20% damage reduction (plus, of course, the no-death thing), while our 4 piece is 10% physical damage reduction that's up 40% of the time. Really its not about damage smoothing anymore, is it? Its about EH as your advice above kinda suggests.

You're discounting the strongest part of AD. The "no-death" thing is pretty much the best part of that spell. The 20% damage redux is nice, and makes it a useful button. The get out of death free card makes it great.

Originally Posted by Wrathblood

This could be a difference in perspectives. In 10-man heroics, I can be spiked down, but my healers can also be overwhelmed. To me, simple damage reduction is less valuable than spike reduction, but it still has value. The way you're describing your stat prioritization, it honestly doesn't sound like Haste OR Mastery is the correct answer. If DR has no value and you're at risk of being burst down every ~3 GCDs by Feng, the correct answer is STAM because you're in Heroic ICC tank spike damage-land.

Edit - In fact, I wonder if there's an argument to be made for stacking STAM over, say, Exp. After all, if you're in 2-swing territory, capping Hit/Exp and stacking Mastery only cuts your chance of a 90% damage spike by about 1/3 over stacking Avoidance. In 3-swing territory, it cuts it about in half. Dumping 7600 itemization points into STAM instead of Hit/Exp would buy you a bunch of EH and I wonder if that might actually be more valuable.

Indeed, which is why I value stamina so highly in the weights I gave to AskMrRobot. But you can't reforge for stam, so there's still value in going for hit/exp cap with reforging. But enchanting and gemming? Yeah, stam wherever you can. I still pick up socket bonuses, but you could choose to forego them for a fight like Feng.

Capstone · 10/29/12, 4:17 AM

Originally Posted by Theck

I tend to use hit/stam in yellow sockets, in order to reforge more into haste. But haste/stam in a yellow socket is also reasonable, especially if the numbers work out such that it lets you reforge to hit-/exp-caps with less wasted overage. Blue slots get stamina, period.

Correct me if I'm wrong, but hit has been blue for quite some time. So exp/stam for red, haste/stam for yellow, and stam for blue sockets is what you would suggest?

a civilian · 10/29/12, 8:03 AM

Originally Posted by Capstone

Correct me if I'm wrong, but hit has been blue for quite some time. So exp/stam for red, haste/stam for yellow, and stam for blue sockets is what you would suggest?

Hit is typically a blue stat but hit/stam (nimble wild jade) exists and is green. (Possibly a bug/oversight, as level 85 nimble gems were dodge/hit instead.)

bromli · 10/31/12, 3:01 AM

Perhaps I'm just doing it wrong, but I cast a fair number of WoGs on other targets. HA during high raid damage is a great spot, depending on other factors. Tank swaps are another. I know it doesn't contribute at all to my personal survival, but I consider it my job to help the other tank survive as well. SS, HoPur, bubble-sac, and off-Wogs help significantly. You could even argue that the healers will have more of their CDs available for when you are tanking. My point is, i think haste is being sold short compared to mastery on tank swap fights, as bastion does nothing for WoGs on others, while haste does contribute.

Charybdis · 11/01/12, 3:39 PM

There are two considerations with that idea: The first is that if your own survival is well in hand then helping out in other areas can definitely contribute to clearing the encounter. Haste is a great way to do that since it gives us more resources over time and thus we can do more in an encounter.

The second consideration is that by going mastery you can trivialize certain encounter mechanics even more and thus allow your healers to focus more of their attention on other things. Depending on your healers they might even decide to get some DPS in either with straight DPS stuff like our own Denounce or with abilities that contribute to both DPS and healing like a Monk's Spinning Crane Kick.

In short, it will mostly depend on how your raid does things as to whether you'll get a better overall benefit from haste or mastery. Of course, if your raid still balks at the idea of a plate tank going haste and thus you don't get that gear then you're probably SoL on that strategy anyway.

Bemxuu · 11/01/12, 8:06 PM

Speaking of Feng and frequent tank swaps... Is it even possible for a Pally to catch all of his double swing with theirs ShotR up? If I'm correct that it isn't, a protection paladin has to use some cooldowns to survive during which he can pool some HoPo once again. So, with no-haste-build tanking him would look like this: 6 sec of ShotR through pooled 5 HoPo, then 12 seconds of cooldowns duration, then 6 seconds with HoPo pooled during that cooldowns, which makes 24 seconds. With high haste (or some luck or, better, both), though, one can keep up his ShotR for 9 seconds pre-cooldowns and 9 another seconds afterwards, which makes perfect 30 seconds of instagib-proof tanking. Here are some rough calculations:
The best ability sequence excluding random AS procs for generating 4 missing HoPo within the shortest number of global cooldowns is: Judge-CS-GCD-GCD-CS-Judge. Timeline for 0 Haste looks like this:
1st GCD: 5 HoPo - 1st ShoR Up - Judge.
2nd GCD: 3 HoPo - 1st ShoR lasting - CS.
3rd GCD: 4 HoPo - 2nd ShoR Up - empty GCD.
4th GCD: 1 HoPo - 2nd ShoR lasting - empty GCD.
5th GCD: 1 HoPo - No Shor Up

- CS.
6th GCD: 2 HoPo - Judge - immediately 3rd ShoR Up.
So, in order to execute that sequence without having that 1,5 seconds of ShoR down, we must "shrink" 6 0-haste GCDs into 5 actual GCDs by means of haste. W/o haste buffs that should be 8500 rating. Probably, below 4000 for buffed state (not really sure, though).
Does anyone see any mistakes here?

rozetta · 11/06/12, 2:45 AM

There seems to be a common train of thought about crit being a non-useful stat for survivability. However, with Seal of Insight + Glyph of Battle Healer, I would assume that melee and yellow crits would proc bigger heals. Given what I've been seeing in parses related to the amount of healing done via SoI and the glyph, is it fair to completely rule out crit as a survivability stat, especially in an environment where a) vengeance is high and b) Tank DPS is a contributing factor to raid DPS?

I can imagine that, as we move up in tiers the relevance of these factors may diminish, but I'm just throwing this out there for the current content.