# Patch 5.2 Tankadin Smoothness Simulations

With 5.2 around the corner, we can feel  pretty certain that the version of Grand Crusader on the PTR is what we’ll be receiving in the patch:

Requires level 50
When you dodge or parry a melee attack, or hit with Crusader Strike or Hammer of the Righteous, you have a 12% chance of refreshing the cooldown on your next Avenger’s Shield and causing it to generate a charge of Holy Power if used within 6 sec.

We’ve already seen that with the 5.0/5.1 version, which is 20% chance on CS/HotR with no avoidance component, our gearing priorities slanted strongly towards control stats: hit, expertise, haste, and mastery.  This ends up being reinforced by a subtle synergy between the control stats – a feedback loop of sorts.  Getting more haste makes mastery, hit, and expertise better, getting more mastery makes haste, hit, and expertise better, and so on.  Subsequently, avoidance just becomes weaker and weaker, to the point that many of us treated it as an afterthought.  Worse yet, in many cases we actively dodged picking up avoidance gear.

We also saw that an earlier PTR version which gave us 0% chance to proc on CS/HotR and 30% chance to proc on avoidance shifted the goalposts somewhat.  Triggering the proc purely from avoidance diminished the value of haste, mastery, hit, and expertise, though not equally.  Haste took the largest hit, and the Control/Mastery set started showing clear dominance over the Control/Haste gear set.  In addition, avoidance started performing pretty well in terms of raw statistics.  There was very little difference in the representation of damage spikes between a Mastery/Avoidance build and a Control/Mastery build.

It was a very interesting region of parameter space, because in most cases you could gear just about however you wanted and still be roughly as effective from a survival standpoint.  I think that for talented tanks, the extra control afforded by Control/Haste would probably still have been their optimal build, because a talented tank shifts their SotR casts forward or backward based on current health levels – something the sim isn’t sophisticated enough to model.  But for your average Joe Tankadin, avoidance gear would’ve been perfectly reasonable.

But it looks like we’ll be receiving this bipartite solution, triggering on both CS/HotR and avoidance at 12% each.  For short, we could call this the “12%/12%” version of Grand Crusader, as compared to the earlier “20%/0%” and “0%/30%” versions.  And we expect that this version will shift some of the mitigation value back out of avoidance (and back into haste).  The only question that remains is exactly how much of an effect it will have.

Luckily, these are the sorts of questions I’m good at.

However, before we fire up the Matlaberator, there’s one more complication we want to deal with.  With a new patch comes new gear, and that may have unforeseen effects on the results.  Whereas in ilvl 496 gear, we were pretty starved for rating after reaching hit and expertise caps, raising the bar to ilvl 522 gives us plenty of extra rating to play with.  It’s entirely possible that all that excess rating will push us into new regions of parameter space.  So to make sure we’re going to be getting results we believe in, we need to update the gear sets we’re using.

Gear sets

The table below contains the stats for the gear sets I’m using this time around.  Each set has 65k armor, 15k strength, and 24150 rating to distribute amongst the secondary stats.  This is roughly equivalent to an average equipped ilvl of 522.

|    Set: |  C/Ha |  C/Ma |  C/Av | C/Bal |    Ha | Avoid | Av/Mas | Mas/Av |
|     Str | 15000 | 15000 | 15000 | 15000 | 15000 | 15000 |  15000 |  15000 |
|   Parry |  1500 |  1500 |  7500 |  4125 |  1500 | 10825 |   7717 |   4000 |
|   Dodge |  1500 |  1500 |  7500 |  4125 |  1500 | 10825 |   7717 |   4000 |
| Mastery |  1500 | 13500 |  1500 |  4125 |  1500 |  1500 |   7716 |  15150 |
|     Hit |  2550 |  2550 |  2550 |  2550 |   500 |   500 |    500 |    500 |
|     Exp |  5100 |  5100 |  5100 |  5100 |   500 |   500 |    500 |    500 |
|   Haste | 12000 |     0 |     0 |  4125 | 18650 |     0 |      0 |      0 |

In previous posts, I maintained a minimum amount of dodge and parry for each gear set to make them more “realistic,” because in practice you often have to make do with a few sub-optimal pieces.  I received a bit of feedback that these minimums were too constraining, partly because players were accumulating true “best-in-slot” sets with nearly zero avoidance, and partly because people were interested in seeing exactly what happened in those extreme cases.  So this time around, I’ve relaxed those constraints significantly.  We enforce a minimum of 1500 dodge, parry, and mastery rating, and only a minimum of 500 hit and expertise rating.  While these keep us from reaching the far extremes, we still have 18650 rating to allocate at will, which is quite a bit – more, in fact, than we had total in the 5.1 simulations.

Most of the gear sets are unchanged in concept.  We still have the standard Control/Haste, Control/Mastery, and Control/Avoidance sets.  I’ve kept Control/Balance, but dropped the Control/Balance variant that was just under hit and expertise cap.  We’ve been over those results before, and it was clear that it offered nothing we didn’t already get by comparing the other gear sets.  In its place, I’ve inserted a “Haste” set that sacrifices all but the minimum hit and expertise amounts in favor of moar haste!  The Avoidance, Avoidance/Mastery, and Mastery/Avoidance sets are more or less allocated the same way as before.

Simulation

This is still using the standard Monte-Carlo code, updated with the 5.2 mechanics.  Just as last time, here’s the copy/paste summary of how the simulation works:

To better understand the data below, here’s a rough overview of how it’s generated: I run a Monte-Carlo sim that simulates 10k minutes of combat (think Simcraft, but paladin-specific and more limited in scope), making all combat rolls and logging all damage events.  I take the resulting string of attacks (something like “1, 0, 0.7, 1, 0, 0, …” where 0 is an avoid (no damage), 1 is a full hit, 0.7 is a block, and so on) and do some calculations on it.  I calculate the average damage intake normalized to 100% possible throughput (i.e. “1, 1, 1, 1, 1, …”), and report that in the “mean” row, representing mean damage intake (lower is better, represents better TDR).  “std” is just the standard deviation of that mean as averaged over 5 attacks.  “S%” is SotR uptime.

The rest of the rows are smoothing data for strings of N attacks.  For now, let’s just consider the first gear column (C/Ha).  I take the damage event sequence and perform a moving average on it (i.e. an N-attack moving average).  I then calculate how many of those N-attack averages exceed a certain threshold of maximum throughput.  So for example, if we look at a 3-attack moving average, the “70%” row tells me how many of those 3-attack averages exceed 70% of the maximum throughput.  Max throughput for 3 attacks is “1, 1, 1″ or 3 normalized damage, so the 70% row tells me how many exceed 2.1 damage.  And so on for 80% and 90%.  Note that they’re cumulative, so if 5% of attacks exceed 90% max throughput those attacks are also being counted in the 80% and 70% rows (thus, if 17% of attacks exceed 80% max throughput, the percentage between 80% and 90% is 17%-5%=12%).  I should add that the repeatability on these simulations is quite good thanks to the long integration time – results usually fluctuate by less than +/- 0.1% (absolute, i.e. 5% +/- 0.1%).

I do this for a bunch of different gear sets, i.e. “C/Ha” for Control/Haste, etc.  The first table lists all of the gear configurations so you get a rough idea of what they look like.  They’re roughly equivalent to stats in ilvl 496 gear.

The code can be found in the matlabadin repository, as usual.  The two files in particular are pally_mc.m and pally_mc_smooth.m.

Data

There is one change here in terms of data presentation.  I’ve found that it was becoming more and more interesting to consider the 60% and 70% thresholds while comparing the different gear sets, so I modified the code to spit out those percentages as well.  Keep in mind, however, that these stats are not always the most relevant measure.  For example, you’re unlikely to care much about the number of attacks that exceed 60% throughput for a string of 2 attacks, because that will be a fairly significant portion of them (you basically need to avoid one of the two attacks or cover both with some combination of blocks and SotR to be below 60%).  On the other hand, if a 5-attack string exceeding 80% throughput is dangerous, a 6- or 7-attack string exceeding 70% or 60% might be dangerous too.

With that out of the way, let’s have at the data:

| Set: |    C/Ha |    C/Ma |    C/Av |   C/Bal |      Ha |   Avoid |  Av/Mas |  Mas/Av |
|   S% |  0.5224 |  0.4106 |  0.4191 |  0.4528 |  0.4990 |  0.3664 |  0.3632 |  0.3577 |
| mean |  0.6013 |  0.5410 |  0.5333 |  0.5559 |  0.6082 |  0.4946 |  0.5036 |  0.5158 |
|  std |  0.1478 |  0.1453 |  0.1759 |  0.1611 |  0.1523 |  0.1879 |  0.1771 |  0.1650 |
| ---- |  ------ |   --- 2 |  Attack |  Moving | Average |  ------ |  ------ |  ------ |
|  60% | 46.4365 | 42.2245 | 36.0532 | 42.0223 | 46.8495 | 30.6360 | 32.6615 | 38.9607 |
|  70% | 40.7703 | 21.9045 | 32.3510 | 32.0562 | 41.5132 | 27.8290 | 27.5790 | 22.6745 |
|  80% | 14.6537 | 21.9045 | 16.8478 | 17.5670 | 17.2325 | 16.7553 | 19.2880 | 22.6745 |
|  90% |  9.8918 | 10.2063 | 11.3503 | 10.7805 | 11.6783 | 11.2970 | 10.6098 | 10.0570 |
| ---- |  ------ |   --- 3 |  Attack |  Moving | Average |  ------ |  ------ |  ------ |
|  60% | 53.8772 | 32.5935 | 39.5635 | 45.3188 | 54.4480 | 34.0340 | 35.6205 | 31.5645 |
|  70% | 37.0875 | 18.1815 | 24.6943 | 28.7108 | 37.0938 | 18.9820 | 17.2312 | 17.7622 |
|  80% | 15.1958 |  7.4453 | 13.2320 | 12.3843 | 16.6545 | 11.3433 | 11.0848 |  8.8108 |
|  90% |  1.2962 |  2.7432 |  3.2145 |  2.6515 |  3.1935 |  3.6222 |  3.2912 |  3.0520 |
| ---- |  ------ |   --- 4 |  Attack |  Moving | Average |  ------ |  ------ |  ------ |
|  60% | 53.9435 | 33.9070 | 39.4125 | 42.8023 | 54.8250 | 32.6215 | 33.1535 | 30.3133 |
|  70% | 32.4230 | 12.4927 | 21.3595 | 25.8540 | 32.6262 | 17.0912 | 17.2385 | 13.6712 |
|  80% |  3.3830 |  5.2843 |  6.2967 |  6.5102 |  6.9973 |  5.6180 |  5.2748 |  6.7035 |
|  90% |  0.0000 |  1.2263 |  0.9175 |  0.6537 |  1.4930 |  1.9408 |  2.3090 |  2.7103 |
| ---- |  ------ |   --- 5 |  Attack |  Moving | Average |  ------ |  ------ |  ------ |
|  60% | 55.5823 | 39.0590 | 40.1595 | 42.7240 | 56.8680 | 33.0050 | 30.1425 | 34.5108 |
|  70% | 28.7618 | 10.5665 | 19.3762 | 22.4620 | 30.4220 | 15.4128 | 13.3293 | 11.5990 |
|  80% |  8.9097 |  1.4918 |  7.0112 |  5.3835 | 10.8432 |  4.9352 |  2.7422 |  3.4403 |
|  90% |  0.0000 |  0.0000 |  0.5200 |  0.3295 |  1.0175 |  1.0540 |  0.6695 |  0.7688 |
| ---- |  ------ |   --- 6 |  Attack |  Moving | Average |  ------ |  ------ |  ------ |
|  60% | 54.3558 | 35.7295 | 34.1108 | 41.2287 | 55.5080 | 25.9360 | 28.3017 | 31.3457 |
|  70% | 24.3050 | 11.0603 | 13.4900 | 16.6937 | 26.1385 | 10.4652 | 10.5652 | 11.1263 |
|  80% |  4.2673 |  0.0000 |  4.3717 |  4.8828 |  6.8168 |  3.4403 |  2.7655 |  1.4698 |
|  90% |  0.0000 |  0.0000 |  0.0000 |  0.0000 |  0.2012 |  0.3240 |  0.2855 |  0.2027 |
| ---- |  ------ |   --- 7 |  Attack |  Moving | Average |  ------ |  ------ |  ------ |
|  60% | 53.2705 | 32.1395 | 32.5230 | 38.7373 | 54.7385 | 24.7587 | 25.7575 | 28.3533 |
|  70% | 22.7233 |  9.7105 | 12.6490 | 15.3603 | 24.6157 |  9.4672 |  9.3260 |  9.6970 |
|  80% |  3.8538 |  0.7848 |  3.0023 |  2.9965 |  5.5295 |  2.0965 |  1.9265 |  1.4485 |
|  90% |  0.0000 |  0.0000 |  0.0000 |  0.0000 |  0.0445 |  0.1020 |  0.1087 |  0.1403 |

The first and easiest comparison to make is between the different Control strategies.  Control/Haste still claims the title for “fastest to null out 90% spikes,” which is achieves for a 4-attack string.  Funny that, haste winning an award for speed.  However, it’s worth noting that Control/Haste consistently underperforms compared to Control/Mastery in just about every other aspect.  It’s generally worse at reducing 80% spikes, in many cases by factors of 3x or more.  And the 60% and 70% data shows us that it falls behind in those categories as well.  Haste is still very good at eliminating that top tier of spikes, but it just can’t keep up with the broad-spectrum smoothing that mastery provides.

This is reflected in the stationary stats as well.  Control/Haste boasts a massive 52% uptime on Shield of the Righteous, but also a rather high mean damage intake level of 60.1%.  By comparison, Control/Mastery drops to about 41% SotR uptime, but also drops mean damage intake to 54.1%.

Control/Avoidance doesn’t do much better than C/Ma in either of the bulk categories – slightly higher SotR uptime, slightly lower mean damage intake, but nothing to write home about.  I don’t have Holy Power generation stats on this table, but you probably won’t be surprised to hear that the extra uptime is directly linked to the extra Holy Power income from having more avoidance.  For smoothing purposes C/Av doesn’t do too poorly, keeping up fairly well for short strings.  However, it starts to fall behind for strings of 5+ attacks.  While it’s not “bad,” per se, there’s also no obvious advantage over C/Ma.

Control/Balance is a weird mix – in some cases it outperforms C/Av, in others it falls behind.  It’s actually about as good as C/Ha in most of the smoothing categories, giving up some ground in certain categories but gaining it back in others.  I’d argue that C/Bal is a better model for the type of gear that we’re likely to be working with as we progress through the Throne of Thunder, as many of us end up using a mixture of “ideal” Control/Haste or Control/Mastery pieces and “less than ideal” dodge/parry tanking items that we end up with because nobody else wants them.  So it’s reassuring to know that even with this patchwork gearing strategy, we’re not suffering in the mitigation department.

The “Pure Haste” build is sort of interesting.  It manages to rack up about 50% SotR uptime, but also has the highest mean damage intake of any set at 60.8%.  While the stats are similar to those of C/Ha, it is uniformly worse in every category.  So while this isn’t actually a terrible gearing strategy, it’s also clear that there’s no significant advantage to doing it instead of going with C/Ha.  It’s worth noting that at all Vengeance levels, hit and expertise are better DPS stats than haste up until the spell hit cap (spoiler alert: this will remain true in 5.2, even though I won’t get around to updating my Maintankadin thread until this weekend at the earliest).  So you shouldn’t be dropping below 7.5% hit and 7.5% expertise to stack more haste if you’re trying to optimize DPS.

Finally, let’s consider the avoidance-heavy builds.  In a surprising turn of events, these really aren’t half bad.  The pure avoidance gear set is actually rather competitive in the 80% and 90% categories.  It’s strictly worse than C/Ma for most purposes, but you could argue that it’s actually stronger for stochastic mitigation than C/Ha.  It’s especially worth noting that for longer strings, it significantly reduces the representation of 60% and 70% spikes compared to C/Ha, which is one of the only areas where 60% and 70% spikes are of interest.  Again, I’d argue that C/Ma tends to do this job better overall, but it’s clear that a “stack avoidance to the sky” strategy isn’t as far behind as it was in 5.1.  It also gets to claim the title of “lowest mean damage intake,” shrinking that value down to 49.5%.

Shifting some of that avoidance to mastery seems to have one major effect: It makes the gear set slightly better at 80%/90% spike mitigation.  We also see a minimal loss of SotR uptime and a small increase in mean damage intake.  70% and 60% spikes fluctuate somewhat as well, but not consistently enough to draw a good rule of thumb – the best we can say is that for long strings, 60% and 70% spikes tend to go up, while for short strings there isn’t a reliable trend to the fluctuations.  All in all, this means that any mixture of mastery and avoidance is going to be about as effective as any other mixture.  Stacking mastery to the sky isn’t going to be that much more effective than stacking avoidance to the sky, at least in the absence of a significant amount of hit and expertise.

Conclusions

While it was nearly impossible to pull a clear winner out of the 0%/30% version of Grand Crusader, the 12%/12% version is a little more clear-cut.  Control/Mastery is the standout for raw mitigation power.  It just does the best job of squelching spikes across multiple categories.

The curious thing about the results is that while C/Ma is definitely the winner, several other gear sets manage to perform rather well.  Control/Haste and Control/Balance are both respectable, and even the gear sets centered around avoidance and mastery have their strong points.  C/Av and pure Haste are about the only two that don’t really seem to have a niche.

All of that said, there’s an important caveat to consider here.  These stat weights are stochastic, and calculated based on a model where the player blindly uses SotR as available.  It does not take into account certain nuances that separate a talented player from a novice.  For example, a talented player may delay SotR if they’re at full health even if they have 3 Holy Power.  If I’m not in imminent danger, I will often sit on that Holy Power until I reach 5 and have a generator coming up in the next GCD, to ensure that I can cast another SotR immediately if it’s needed.

That sort of nuanced play has a peculiar effect, in that it’s essentially weighting the dice in your favor.  You don’t need SotR coverage when you’re sitting at full health and have just avoided an attack, so you’re subtly shifting that coverage to periods of higher risk.  That has a small but noticeable effect on mitigation stat weights, giving a slight edge to haste-based control strategies.  This is a topic of frequent conversation at Sacred Duty Headquarters, with it normally playing out something like this:

Theck: “But the numbers say that C/Ma is simply stronger based on stochastic metrics”
Meloree: “Yes, but your numbers are shit, because nobody cares about stochastic metrics.”
Theck: “You’re mean.”

I’ve highly paraphrased and villain-ified Mel there, but I have trouble disagreeing with the core message he’s trying to get across.  Not that my numbers are shit, but that a stochastic model has its limits.  Given the option between high uptime on a moderate-mitigation buff (i.e. C/Ha) and lower uptime on a higher-mitigation buff (i.e. C/Ma), the higher uptime version is often safer.  Not because it does better in stochastic “how much damage did I take” metrics – in fact, quite the opposite, it usually loses those comparisons.  But because you’re not in as much danger during a high-mitigation phase, such as having SotR up.  Making that high-mitigation phase “safer” isn’t as valuable as having more of those phases in the first place.  If you wanted, you could include healer mentality in that argument as well.  A healer will respond differently to a sequence of “Big Hit, little hit” than to a sequence of “moderate hit, moderate hit.”

So, what does that mean when we apply it to our data?  Well, Control/Mastery wins by the numbers, but I think that Control/Haste probably still wins by Meloree fiat.  You give up some extra time-averaged mitigation and open yourself up to more spikes if you play poorly.  But if you play well and time-shift SotR casts according to current health and other dynamic encounter variables, it’s very likely that Control/Haste will meet or exceed the survivability level afforded by Control/Mastery.  Remember, you’re getting about 10% more SotR uptime (absolute, a 25% relative increase), and if you play well you’re applying that extra uptime to the sections of the fight you need it most.  And, lest we forget, there is a significant DPS gain to be had by going with Control/Haste, which is a big factor to consider for heroic progression, especially in 10-man content.

It’s almost a shame that the 0%/30% version of Grand Crusader didn’t stick.  In that version, there was a significant mitigation cost to going with Control/Haste instead of a more “tanky” gearing strategy, which seems like a reasonable trade-off for the DPS gains.  Even though it would have reversed our gearing paradigm, it would have made the gearing paradigm a bit more logical.

Instead, we get the 12%/12% version, which basically maintains the status quo.  Control/Haste is no longer the clear leader as it was in 5.1, but it’s probably still the best strategy for a progression-oriented tank.  Control/Mastery is an equally viable choice, and probably a better choice for intermediate tanks that aren’t as careful with their SotR timing.  Avoidance builds lag slightly less than they did previously, and are still a good choice for beginner tanks that are still getting comfortable with active mitigation.

TLDR:

Advanced tanks will gear for Control/Haste or Control/Mastery, probably depending on encounter specifics and gear availability. Your BiS gear will probably involve a lot of hit/haste and expertise/haste retribution gear.

Intermediate tanks will probably stick with Control/Mastery for better stochastic mitigation. For these tanks, there’s probably no need to chase haste gear.

Beginner tanks will likely stick with traditional tank gear that’s high on avoidance and mastery.

This entry was posted in Tanking, Theck's Pounding Headaches, Theorycrafting and tagged , , , , , , , , , , , . Bookmark the permalink.

### 57 Responses to Patch 5.2 Tankadin Smoothness Simulations

1. blaine4848 says:

Great stuff Theck. Thanks again for all the hard work in breaking this down.

2. Hamsda says:

3. Newsom says:

You have no idea how much I appreciate you doing these. Just in time for 5.2 too. Thanks!

4. Dan says:

My guess is this is Blizz’s way of easing it in, to avoid people having to massively refarm gear as a tier drops….I expect mid 5.2 for a patch to shift the numbers a little, and an early patch 5.3 note that shifts it further…

• Theck says:

I had that thought as well. It wouldn’t be hard for them to shift it over time to make haste gear a less attractive option. And it would give them the opportunity to announce it well in advance, so players can plan their gear choices accordingly.

That said, I’m not sure they will do that. There was a lot of support for the haste paradigm from the paladin community on account of it being fun. The real question is whether it’s worth maintaining that despite the in-game limitations (haste not being on prot loot tables, competition with plate DPS specs), especially given that none of the other plate tanks value haste very highly.

• Lakh says:

Avoidance has never been an eye-popping success – it’s just not visceral enough. I’d much rather see them bring the other tanks in-line with the fun of haste, and I’ve got a sneaking suspicion that might be one of the schools of thought at Bliz too.

TBH after seeing haste in action, I’d be perfectly happy if they just did away with tanking plate & mixed all the strength plate in together.

• Meloree says:

Before DR on avoidance, when trinkets like Moroes Pocketwatch added 18% flat avoidance and tanks were easily able to get over 70% pure avoidance from gear, avoidance was very powerful, and very useful.

• bryjoered says:

Avoidance was really good in cataclysm too. I like that tanks have their own classification of gear, it eliminates the possibility of other people rolling on it. Just like healer gear has spriit, and each dps spec has their own gear, it just makes sense doesn’t it? I never steal dps gear from melee unless they don’t want it it just works, it’s easy, and honestly isn’t gonna kill your performance as long as you reforge and gem correctly.

• Theck says:

Were we playing the same Cataclsym expansion? Because avoidance was still pretty crap in mine. Mastery and Stamina were both good, but everything else was weak. Avoidance just happened to be “crap rising to the top of a pool of crap,” in Mel’s words.

• Kameron says:

Theck I have wierd sense that is telling me after reading all this that haste/avoidance might not be a terrible combination and exactly the synergy I’m looking for. At early raiding avoidance gear seems to drop like mad. Is this a viable gear strategy for a young blossoming tankadin. Make use with the best you have ? It seems blizzard are clearly getting away from gains using complex analysis such as your own and making all things workable to an extent and w/ skill haste and avoidance could facilitate favorable outcomes in older raid content.
.

5. Duncan says:

Spending what little free time you have on this is most appreciated T!

QQ: I’ve already been using SotR as more of a “cooldown” button, when I know there’s an exceptional amount of damage incoming. Should I keep it up more often (yes, that is what she said), perhaps holding off and going in to “cooldown” mode when needed?

Again, thanks!

-Duncan

p.s. You should create a donations button.

• Theck says:

There’s really no point in holding SotR if you’re maxed out on Holy Power. Using the example I gave in the post (and in a comment below): let’s say I bank up 5 holy power and my next GCD is going to be a Crusader Strike. Is there any advantage to holding SotR at that point and wasting the Holy Power generated by that CS? I’d argue no. If I just cast SotR, I 1) get 3 seconds of relative safety, and 2) that next GCD brings me back up to 3 Holy Power, so I can cast SotR again immediately if I need to.

Given how rapidly we generate Holy Power, it’s sort of silly to sit on SotR until we take a big spike. In most cases, it’s not worth sitting at 5 Holy Power “just in case,” because if you do that for more than about 5 or 6 seconds you basically wasted an entire SotR anyway, so you may as well have cast it to begin with.

I’ve thought about a donations button, or adding ads to the blog, but I’m not sure it’s worth it. I can’t imagine many people would care to donate, and I don’t think google ads would generate much revenue on a site that only gets ~1k hits per day.

• Duncan says:

My initial comment was badly worded.

I don’t leave it alone the entire duration of the fight, but I do *hold back* a bit when some mechanic is about to hit. Now however, I’ll keep it on CD.

In lieu of the button, do know your work is appreciated.

Again, many thanks.

• Theck says:

Yeah, holding it for a GCD or two to line it up with a boss attack is often optimal (that’s the “time-shifting” argument). Firing it off as soon as you have 3 Holy Power isn’t going to be as effective as banking and spending wisely.

But as I noted, it’s almost never going to be worth sitting on it if you’re at 5 HP *and* your next GCD is a holy power generator. Because you could cast it (off-GCD), get your 3 seconds of coverage, and be back at 3 holy power before the buff drops off again anyway.

6. “If I’m not in imminent danger, I will often sit on that Holy Power until I reach 5 and have a generator coming up in the next GCD, to ensure that I can cast another SotR immediately if it’s needed.”

Question: Don’t you risk losing Abalaster Shield, and Bastion of Glory stacks if you delay your SotR?

• Theck says:

That’s not really a problem, no. You generate ~0.4 holy power per second in current gear (higher if you stack a lot of haste, HPG is basically your SotR uptime). Thus, you generate 5 holy power in about 12.5 seconds on average, assuming you use a Control strategy. We’re not talking about sitting on SotR any longer than that – that would be wasting potential Holy Power. If you have 5 Holy Power and a generator coming off of cooldown, there’s no reason to delay SotR any further.

Bastion of Glory lasts 20 seconds, so you’re basically guaranteed to refresh that before it expires.

The Alabaster Shield buff lasts 12 seconds. Remember that casting SotR consumes *all* stacks of Alabaster Shield, so any charges that existed *before* that 12.5-second period between SotR casts are already gone. The chance that a block occurs during the first second or so after your previous SotR cast is pretty small, so new stacks will generally last until your next SotR. And with ~20% avoidance and ~30% block, you have a fairly good chance of getting more than one block in 12.5 seconds (~14.7% assuming 8 attacks in that period). So you generally won’t see wasted Alabaster Shield charges either.

7. Tim says:

I disagree with the order of advanced verse intermediate gearing strategies, though perhaps this may be to me not having a full understanding of it. It would seem to me that someone who understands those nuances of tanking (holding HP for the most valuable times) would benefit the most from the c/mastery strategy. Take for example Sha of Fear (I’ve only done it normal to this point so I’m really looking at just that part). That fight was a fight where mastery was extremely dominant, which was in large part due to the Thrash mechanic. If you’re able to use SotR during the most important times (right before thrash) then even if you cut down on overall uptime you are actually making your uptime much more effective. The benefit of the c/haste build to me seems to be its ability to have a higher uptime on SotR which seems like it would be more beneficial for an intermediate tank who may not use it at the most effective times. Obviously I’m discounting the dps contribution of c/haste but that really is more of a secondary affect than a primary gearing strategy. Is there something I am missing here?

• Theck says:

Sha of Fear is a bad example to base that conclusion on. It’s a bit of an outlier, because it’s an encounter which gives you very large damage spikes at predictable, widely-spaced intervals. On heroic mode it is very common to use exactly the gearing strategy you suggested – stack mastery to the sky, because haste does you little good. You want as much survivability in a small window as possible, and that window has a very short “duty cycle” (to use electrical engineering terms).

However, that encounter is the exception, not the rule. If you’re trying to smooth out spikes in throughput damage, then you want a model where the boss attacks relatively quickly so that the threat is more or less constantly there. While most bosses have special abilities that the tanks do need to survive, they’re generally not so dangerous that the baseline 38% mitigation of SotR isn’t already a sufficient cooldown.

So, if you’re trying to reduce the number of spikes you take from a boss that hits very quickly, what’s better: having short periods of high mitigation with large gaps, or having short periods of moderate mitigation with much shorter gaps? Taken to an extreme: would you rather take 10% damage for 1 second every 10 seconds, or 50% damage for 1 second every 3 seconds? The latter is going to be far smoother.

So the skilled tank can do well with either C/Ha or C/Ma, but C/Ha may give them more flexibility. They get about 10% more uptime on SotR (50% compared to 40% in C/Ma), and by using it intelligently they can apply that uptime more efficiently (i.e. don’t use it if at full health with lots of absorption effects active). Perhaps more importantly, the intermediate tank has no hope of pulling this off effectively, so for them the Control/Mastery scheme is going to be much more reliable. The advanced tank is pretty much the *only* person who can pull off this time-shifting trick.

And to be honest, I’m not sure how high of a skill threshold we’re talking about here. while I consider myself an “advanced” tank, I’m not sure that I’m even capable of pulling that off very effectively during a complicated boss encounter. Doing it for, say, Patchwerk is pretty straighforward. But once I have to start moving a boss around, tracking timers and debuffs, watching the raid to cast utility spells, or what not, my attention gets stretched pretty thin. I certainly don’t think I’d be making most efficient usage of that flexibility on a fight with a lot of moving parts.

So there’s an argument to be made that C/Ma may be a more practical choice even for the advanced tank depending on the encounter in question, which is why I sort of left that category up in the air. And, of course, there’s the DPS contribution to consider. In a 10-man raid, it can be a rather significant deciding factor.

• Meloree says:

The simulations are generic, and the gearing advice produced most certainly doesn’t override the principle of adjusting your gear for specific bosses.

Sha and Thrash are clearly examples where smoothness isn’t nearly as important as “Mitigate as much as possible for Thrash”, and gearing for that fight should surely reflect that. That absolutely doesn’t invalidate sims or the generic advice.

Tuning gear fight by fight is still an important part of giving yourself an edge in progression encounters – generic advice from the sims most certainly doesn’t apply to every boss encounter. The hope is that it applies well to the majority (or even plurality) of them.

8. Wrathblood says:

Nice work as usual, Theck.

One point that occurs to me is tanking in an AoE situation. The coming tier looks like it has an abundance of add fights and I’m going to be tanking with a Monk so you’d think he’d be the one picking them up (Keg Smash makes me sad sometimes). However, he’s still fairly new to tanking and we haven’t completely ironed out our plans yet.

At first glance, I’d imagine that adding more adds (i.e. dramatically reducing the swing speed you face during an encounter) would significantly improve the performance of the Avoidance-driven sets. Obviously there aren’t any AoE threat tools available via HoPo but at the very least it would somewhat increase your ShoR uptime. As such, I’m wondering if the C/AV set might have its niche as the aoe-tanking set.

• Theck says:

Technically AS could be considered an AoE threat tool if you leave it unglyphed. But I agree, in an AoE situation avoidance should give a pretty good showing. It’s an open question whether C/Av will have an advantage over raw Av there – since AS doesn’t need to hit to give holy power, the control aspect is somewhat diminished if you’re getting hit frequently enough to spam AS. I have to update the MATLAB thread before I get to run the AoE defensive sims, however.

• Duncan says:

Does this mean we might have an optimal build for off-tanking (adds) and main-tanking (boss)?

Since you’re in bed with Mr. Robot (or Ms. Robot), would there be an “AoE” build, or should one assume that the C/Av is just that?

• Theck says:

I think that if you’re tanking 2-3 mobs, it’s probably not worth completely re-gearing. If there’s a fight where you’re tanking more than that (heroic Shek’zeer for example), then it would probably be worth switching to an avoidance- and mastery-heavy set. I’d still use C/Av over M/Av or Av, personally, but I really value the ability to plan out my next few GCDs, which you lose once you can start missing.

Note that some adds aren’t L93 though, so you might be able to relax the hit/exp requirements in such a set.

9. Vayacondios says:

Good stuff, thanks much. And thank you for the TL;DR!!!

10. Scoutyou says:

As usual, great work and interesting read.

What I feel should be noted though is that mastery actually have a lot higher value, especially in 10 man.

In 10 man, boss melee swings alone won’t kill you, ever, ever.
When you die it is from a predictable damage spike, the boss casts an attack that deals high damage or whatever, adds spawning etc etc. Something makes the damage skyrocket basically.

For this, mastery have an inflated value since if you can predict the spike you can bank up holy power for it use SotR for it.

If a boss hits for 100k normally and sometimes hit for 600k, I would rather have 55% damage reduction instead of 45% damage reduction on the 600k hit instead of having 10% higher sotr uptime on the 100k hits.

Mastery just wins hands down to survivability.

• Runik says:

Just an appreciation reply, once again a great post

• Theck says:

I don’t tank 10N enough to confirm or deny your claims, but I suspect that there’s at least a bit of hyperbole there. Bosses do hit for significantly less on 10- than on 25-, that much is true. I don’t think it’s fair to say you’ll “never” die from melee alone though. For example, even a 600k hit is survivable, as most tanks will have around 650k health even without stam stacking (I have over 800k). So it’s not just the 600k hit that matters, but the events surrounding it. Your likely death scenario is probably that 600k hit combined with a few leading or trailing 100k hits, in addition to a bunch of healing and absorbs.

The other thing to remember is that if we’re getting hit for 600k, it’s probably a significant event we’re planning for. And that means COOLDOWNS. Part of the reason I model steady-state throughput damage is because ever since Wrath, tank design has centered around us using cooldowns to mitigation predictable (and in some cases unpredictable) spikes. So in my mind, it’s unreasonable to say that it’s the 600k hit that matters, because if I have any worth as a tank I’m already running a cooldown for it, which means it’s no longer dangerous. At that point, the balance shifts back to haste, because I may be able to keep SotR up for more of those surrounding melees.

The exception is a case like Sha, which might hit you for 600k so frequently that you can’t prescribe a cooldown for each one. And that’s exactly the situation where mastery thrives, as you suggest. But I don’t think it’s fair to say that that situation is very general – it’s the exception, not the rule.

• Scoutyou says:

On most heroic 10 boss fights the paladin tank can almost keep himself up with his SoI alone when the boss is only auto attacking. Throw in some small hot or a disc shield and your topped with SoI. Boss melee hits will not bring you low before that huge hit. Also, with good planning you can have sotr up for 5 seconds atleast before the huge hit to prepare for it so that you are 100% topped of in advance.

11. Kimina says:

Theck, thanks again for your work.

Since Dodge and Parry are becoming “functional” stats for us in 5.2 would that increase the value of trinkets like Vial of Dragon’s Blood? I know most agree EH is a good path to follow, but would the static Mastery combined with an EH/Stam trinket be a good option? Not to mention Grand Crusader proc’s from Dodge.

• Theck says:

It will certainly be more attractive than it was in the past. Especially in 10-mans, and maybe 25N now that gear levels are higher. I’m still of the opinion that stamina is the best stat for 25H.

Curiously, the next tier only has one stamina trinket, so we may all be forced to choose a second mastery trinket (or use a ilvl 497 double-upgraded Lao-Chin trinket as our second stam trinket).

12. Runik says:

Just an appreciation reply, once again a great post

PS: sorry for the double pots but replied on Scoutyou reply

13. Barnicus says:

Hi Theck. I’m in the same boat as probably a lot of players in WoW, but not many commenters here on your blog or on Maintankadin. Basically:
I am what you describe as an “Intermediate”. I know my rotation, pull it off with skill, pay attention to strats and am a leader, but don’t have the skill to do things like you describe with timing my AM precisely.

However, I am in what could probably be described as a pretty lower-intermediate or maybe even “skilled beginner” guild. Basically we decided partway through Cata that raiding with our significant others and friends was more important than progression, so now we struggle with normal modes… and are fine with that. This worked out better for us in Cata where we were able to beat all normal mode content before significant nerfs. In Pandaland we are struggling a bit, though, with the exact same raid team. We have only downed half the available bosses on normal in 5.0/.1. We will likely polish the rest of this tier off in normal with the 10% nerf in the next few weeks.

The end result is that I have a very small pool of non-LFR gear to “fish” from each patch. I often have to take what I get. Sometimes that means haste gear, sometimes that means mastery gear. Sometimes that means parry/dodge gear, which makes me sad, but oh well. My question, based on this fact of my raiding life, is: is there anything in the numbers that makes it a bad strategy to mix/match haste and mastery gear? Basically, I won’t be able to stack either, because the drops I get are only part of the overall possible drops due to the fact that we kill/farm only a percentage of the available bosses each tier. I’m sure when we get to Throne we’ll eventually be getting the first 3 to 5 bosses down, but will likely get stuck after that until a nerf, so I’ll only have access to whatever loot the first 3 to 5 bosses have (on normal of course, not counting LFR). That means I might have access to a mastery chest but not a haste chest, or a haste trinket but not a mastery trinket, etc..

Sorry, huge post got away from me. For a guy who only will ever be tanking 10N, is it ok to mix/match Mastery and Haste, given that both are good? Or is it important to pick one or the other? In other words, are they devalued if you don’t stack one? (As a note, “sure, do whatever you want, 10N is easy so it doesnt matter” is not a good answer. Obviously with the raid team I’m on 10N is not easy for us, so gearing for effect to help the team’s DPS and to help my healers DOES matter.)

Thanks. 😛 On a side note, love all your maths and blog posts Theck. Gives a guy something to do while his guild is stuck, keeping me from going mad. 😉

• Theck says:

There’s nothing wrong with mixing and matching. In practice that’s what most of us have to do anyway, even if we’re working on heroics. When new hit/haste and exp/haste pieces drop, I don’t take them over a Ret main-spec, simply because the raid benefits more from the gear going to that player. There’s a lot of dodge and parry items that nobody else can use, so I make do with those until the Ret players get their stuff (I’d add DK’s/Warrs, but ours seem to avoid haste pieces, so maybe haste is a low priority for them).

That’s why I’ve simmed the Control/Balance gear set. It more accurately reflects the average stats of someone having to “make do” with a mix of tanking items and haste pieces. And as you can see from the data and my analysis, it’s not significantly worse than C/Ha or C/Ma. It falls pretty squarely between them, so it’s a perfectly viable gearing strategy, especially if your options are limited.

To put it another way: The fact that it may not be ideal doesn’t make it bad. It’s still a good strategy, it just may not give you quite as much survivability as an ideal set. But it’s generally going to be better to use a higher-ilvl mastery/dodge or mastery/parry item than a lower-ilvl hit/haste item, provided the ilvl gap is significant (i.e. a full tier or so).

Third analogy: the house you would buy if you had a million dollars is likely far different than the place you actually live (unless you’re a millionaire already, in which case.. pretend you aren’t). That doesn’t make your current home bad just because there’s a better one out there – it can still be the best choice for you under the circumstances.

14. Kostas says:

Hey Theck, great read as always, thanks for the info.
I have a question. How is Mastery and / or Haste affected by our t15 set bonuses? Seeing that wog / eternal flame will increase our block chance by a considerable amount would that affect somehow one stat over the other?

Effectively won’t we have to keep up the wog buff + ss + have enough hp to drop those shields on top? Does that make haste stronger in any way?

Lastly, in a fight like will heroic, or in any fight that we have to taunt every 30-40 seconds and has some heavy tank damage basically we will be getting free holy power very often. Will haste pull ahead (since we will have greater coverage with our sotr) or mastery?

Thanks for taking the time to read this.

• Theck says:

The 2-piece bonus isn’t worth treating as a maintenance buff. In other words, you won’t be using WoG every 15 seconds to keep it active; as such, it’s not something we’d want to optimize around (see http://sacredduty.net/2013/01/24/protection-tier-15-set-bonuses/ for more info on that). Either way, it shouldn’t have a large effect on either haste or mastery while it’s active.

The 4-piece will make Divine Protection a more potent cooldown by giving us larger SotR uptime. In some sense, it may be like an extra Holy Avenger, giving us nearly complete uptime on the buff. For survivability during that window, mastery would obviously be superior (because there are fewer gaps in SotR coverage to fill). However, haste will still give you faster HPG, and since SotR is additive it will extend that coverage period and make it longer. So neither one is wasted, and which you prefer mostly depends on whether you’d rather extend that coverage period a little longer or have higher mitigation during the (slightly shorter) period.

15. fetznschaedl says:

Dear Theck,
(I would have sent you the following via PM, if I’d known how to. Sorry if I’m cluttering this board, feel free to delete the post if you deem it too off-topic)
I’d been playing a Tankadin since TBC, and ever since late TBC/the beginning of Wrath, I’ve followed and enjoyed all your free theorycraft publications. Not only are they thoroughly researched, tested and thought through, they are also presented in a fashion that a statistical nullhead like myself can not only understand and appreciate, but also enjoy. Spiced with just the right mixture of data and numbers, sexy graphs and humor, they’ve always been a pleasure to read and digest. I want to express my heartfelt gratitude for your work, which has really helped me a lot over my years of playing WOW (and you don’t even get paid for it, right?).
I know that not only are you an avid theorycrafter (which has to take up quite a lot of time on itself), but you are also a scientist, progression raider (even leader?), AND have a fiancee. All of that surely requires extreme measures in the way of time management (and probably sleep deprivation management, too), and therefore, I normally wouldn’t ask such a thing of you, but (here comes the desperate plea part of my post) would it be too much to ask (or too complicated, or what would you need) to run your sims for DK tanks too, or could you hazard an educated guess as to what stat priorities might be? I did a lot of searching, but all DK theorycraft seems to be based on anecdotal experience and/or “napkin math”, feeling, “I do it cause it seems logical” etc.
Anyway, if this is not easily done, just forget I asked, and I really, really want to thank you again for all the work you do in the community. I think somebody proposed a “Donate” button, I would gladly press it if it were there.
Best regards!

• Theck says:

I would love to run math for other tanking classes in WoW, but there’s only so much time in the day, and as you noted, I already have a lot of responsibilities. It wouldn’t be too hard to write a DK version of the simulation, but it would take quite a bit of time. And I’d need up-to-date information on exactly how a DK makes spell choices nowadays, as the DK resource/finisher systems are more tightly coupled than paladins or warriors are. I could probably get a lot of that information from SimC and EJ, but again, it’s all a matter of having the time to do it. And there’s already a pretty big backlog in the queue of “things to do for paladins,” unfortunately.

But here’s what I would need to make it reasonable to add to the queue:
1) Spell cast priority – i.e. what decisions you make while tanking, in the form of conditional statements. Like “IF (runes available) AND (damage taken in last 3 attacks > X) THEN (cast Death Strike),” etc. Note that conditionals based on player health aren’t possible in this sim, which is a particularly annoying limitation for DKs because their active mitigation is so reactive.
2) Spell details – cooldowns, rune costs, etc. for each spell involved. I’d also need the formula for healing/absorption from Death Strike and Blood Shield.
3) Any other mechanics I should be aware of that modify these effects.

Even with all of that, there’s no guarantee that I’d get to it any time soon. As of right now, it’s unlikely that I’d get to it before summer at the earliest.

• fetznschaedl says:

Thanks for taking the time to answer me. If it’s ok with you, I’ll check and see if I can fit everything required together; that’s gonna take take some time anyway. But if you feel you just don’t have the time/energy/it’s not relevant or interesting to you or if you’re afraid everyone and their uncle’ll come running to you begging for an analysis if you start this, just tell me, no problem at all – I don’t want to take advantage of your altruistic personality. Worst case, I’d just level an alt tank

16. queldan says:

Impressive work as usual Theck. I’ve been mulling over your post for a couple of days, and I’m going to go on a limb and try to paraphrase what I’m reading:

This time around, it’s more important to get good loot from higher ilvl and reforge from there to meet your targets, than to seek specific pieces due to its secondary stats. IF you can get a cool piece with just the stats for you, so much the better, but there a no “wrong” choices, apart from crit pieces.

• Scoutyou says:

Crit is honestly not that bad stat for 10 man tanks, as the dps is just so valuable. Given haste is more valuable for dps but a lot of 10 man tanks prefer crit over dodge or parry. Will see how that change for the next patch depending on the dps value of dodge and parry vs crit.

• Theck says:

Dodge and parry are still terrible DPS stats. As in, less effective than Intellect, which gives you spell crit. Updated DPS sims will be posted on Maintankadin later today or this evening (meaning, as soon as I get to it).

All in all, I think Queldan’s summary is reasonably good. Crit pieces are an odd case, as they’re a straight “DPS vs. survivability” choice, so it’s hard to quantify that.

17. Schroom says:

I’m pretty bugged that the GC change isn’t mentioned in the current Patchnotes…. this is a really big change and should be found there… or yould it be that they change it back for the patchrelease and work on somthing different for 5.3?

• Theck says:

It’s probably just an oversight in the patch notes, but we’ll know for sure tomorrow. I would imagine the spell diff on MMO-Champion would also reveal the truth. It seems awfully unlikely that they’d have reverted it without any comment though, so I doubt it will still be 20%/20% tomorrow.

18. Peter says:

I found out yesterday that my retired-since-wrath paladin gets to tank my main raid next patch… so I’m scrambling to gear him up as quickly as possible. Do you think the new 12% avoidance proc on Grand Cruasder makes River’s Song worth taking over Colossus overall? I’ve seen that you’ve maintained Colossus, presumably on the paradigm that a shield can be counted on and a dodge can’t, but would River’s Song now factor significantly into the smoothing value? Cost is no object.

• Theck says:

The amount of avoidance you get from the proc is relatively small, and you only get 12% of *that* in AS procs, so it’s not really enough to make it much more attractive than it was. Also, River’s Song is dodge – you’d probably be better off using Dancing Steel, which will likely give you a larger amount of avoidance via parry.

That said, if you like Dancing Steel for its extra DPS contribution, I wouldn’t say it’s a bad choice over Colossus. It’s just probably not the most defensive choice, but neither is “bad.”

• Brandice says:

Dancing Steel got a proc chance buff, too (patch notes).

• Theck says:

I hadn’t noticed that, thanks. I may be tempted to switch for the extra DPS myself, especially in challenge modes.

19. Pingback: Control Shift En-tier | Sacred Duty

20. Pingback: Is Nothing Sacred? | Sacred Duty

21. Pingback: Sacréd bleu! | Sacred Duty

22. Pingback: The Case for Haste | ShotRdin