MATLAB DPS Simulations for 5.0.5 and Q & A

I’ve spent the bulk of this week putting some finishing touches on the MATLAB DPS simulations, and I’ve posted the bulk of the results in the 5.x maintankadin thread.  If you want the nitty-gritty details, you’ll have to go read the thread – I’m not going to repost all of it here.  What I do want to do is highlight some of the interesting conclusions we can draw from the results.

In addition, with Mists and level 90 play around the corner, I know a lot of players have questions about mechanics or strategies.  Certainly there are some excellent resources available already – this blog, my MATLAB thread, and my 5.x Talent Guide for starters, but there’s also Rhidach’s 5.0.4 guide, the EJ thread, and the esteemed Knaughty is putting together one of his famous “Let me FAQ you“-s.  I could go on.  However, sometimes you just have an oddball question that you can’t find an answer to in any of the mainstream resources.

To help with that, I want to host a small Q&A style discussion in the comments.  If you have any questions about anything related to paladin tanking, feel free to post that question in the comments.  I’ll try and reply as often as I’m able this weekend, and answer whatever I can.  If there are particularly good questions, or ones that require a longer answer, I’ll append them to the bottom of this post.

MATLAB DPS Simulation Summary

Note that all of these results apply to level 90.  YMMV at 85.  The stats are based on a normal-mode raiding gear set (ilvl 496) with 4-piece protection tier, and all of the numbers I’ll present in this post assume 100k Vengeance AP.

First, here’s a plot of ability damage for our core rotational skills.  :

Crusader Strike and Hammer of the Righteous are both pretty wimpy compared to the rest of our skills.  Curiously, Consecration and Hammer of Wrath are our two highest-damage abilities.  If we glyph Avenger’s Shield (Glyph of Focused Shield), it sneaks ahead of both by a small margin.  This tells us that both HoW and Cons will be important parts of our single-target DPS.

In fact, if we look at the rotation simulation, we find that Cons is so good that it’s a DPS increase to prioritize it over glyphed AS.  Our max DPS rotation (assuming 7.5% hit and 7.5% expertise) is actually:

CS>J>Cons>ES>AS>HW>SotR.

The interesting part is that we only sacrifice about half a percent of SotR uptime by pushing Cons ahead of Avenger’s Shield.  We don’t gain any DPS by pushing Cons higher in the queue, so there’s no advantage to doing so.

Of course, the DPS gain of Cons>AS isn’t that huge either.  We only gain about 150 DPS (out of 96k) by choosing that prioritization instead of

CS>J>AS+>Cons>ES>AS>HW>SotR,

which sacrifices none of our holy power generation (“AS+” means “AS only if Grand Crusader is active).  That said, if we’re concerned about survivability, we’re also going to be keeping Sacred Shield active and applying Weakened Blows.  There are a number of valid priority queues that accomplish this, but the most balanced in terms of survivability seems to be:

^WB>CS>J>AS>Cons>SS[buffSS<5]>HW>SotR.

To clarify, “^WB” means “reapply Weakened Blows if it isn’t present,” and “SS[buffSS<5]” means “refresh Sacred Shield if the remaining duration is less than 5 seconds.”  Putting Sacred Shield at this place in the queue costs us about 1k DPS for about 10k absorption per second and 93% Sacred Shield uptime, which is a fair trade.  Bumping SS up any higher costs more DPS for only a small increase in coverage.  Bumping it down below Holy Wrath reduces the DPS cost to 0, but also drops the uptime to about 74%.  Your level 90 talent would slip in above or below Sacred Shield without having too large an effect on either DPS or SS uptime.

One final note about rotations: Hammer of Wrath hits like a truck, so it’s a DPS increase to put it at the front of the queue during Execute range.  Your priority for max DPS will literally become HoW>CS>(everything else in whatever order you were using).  However, note that doing so will put a dent in your holy power generation, SotR uptime, and SS uptime.  As such, it’s hard to gauge exactly where to put HoW in the priority queue, because it’s up to each player to decide how much survivability they’re willing to sacrifice for it.  The queues listed in the rotation sim post can help you get a feel for the magnitude of that trade-off and determine your comfort level.

Note that if there’s a queue that I’ve omitted and you’re curious about, just ask and I’ll add it.  The simulation takes very little time to run, and is versatile enough to handle all sorts of crazy conditionals.

The talent simulation gives us a few pieces of useful information.  It shows pretty clearly how inferior Eternal Flame is to Sacred Shield.  You sacrifice around 4k DPS and 10% SotR uptime to gain ~16k healing per second by using Eternal Flame as a maintenance HoT.  By comparison, Sacred Shield costs less than 1k DPS and has no impact on SotR uptime, and gives you 10k absorption per second.

It also shows how the L75 talents shake out.  Sanctified Wrath is the clear winner for DPS, so if you’re trying to parse on the MoP version of Baleroc or Ultraxion, that’s the go-to choice.  Holy Avenger takes second place in the DPS competition, and Divine Purpose trails behind.

For survivability, it’s a mixed bag.  Divine Purpose gives you the highest time-averaged SotR uptime, but it’s not very controllable.  Holy Avenger has the advantage of giving you an extra cooldown, during which you can have 100% uptime on SotR, but the time-averaged uptime is a little lower to compensate.  Sanctified Wrath gives you the lowest time-averaged uptime, but also gives you the 20% healing taken buff, which is hard to quantify with this sort of sim.  You do get a higher SotR uptime during Avenging Wrath, which helps, but it requires a rotation change to make full use of it, which could be complicated if you’re not completely comfortable with your rotation yet.

All in all, I think Holy Avenger is the best overall choice for the same reason that I liked Holy Shield’s active version better than its passive version in Cataclysm – control.  But none of these talents are terrible, so you can’t go wrong choosing any of them.  As always, if an encounter mechanic dictates that one is the clearly superior choice, don’t hesitate to swap to it.

The L90 talents are pretty boring.  Execution Sentence is the highest DPS increase, but it’s less versatile in that it’s not that useful as a survivability cooldown.  Generally if you want a large healing burst, you want it now, not 12 seconds from now.  Holy Prism is more versatile, and can be self-cast for AoE damage on many mobs, but its shorter cooldown means it costs more GCDs to use effectively.  Light’s Hammer is a bit of an oddball, because two self-cast Holy Prisms will do more damage in an AoE scenario.  So Light’s Hammer is only a niche ability for DPS (as the targets would have to live less than 20 seconds).  What makes it interesting is the healing aspect of it – it’s a good bit of healing, basically a mini-Tranquility.  We might end up taking it not for DPS considerations, but to have another miniature raid cooldown.  I’d say the default choice is Execution Sentence, but I can see each of the talents being chosen for specific encounters.

The glyph simulation clearly puts Alabaster Shield and Focused Shield out front as our top two single-target DPS glyphs.  Glyph of WoG isn’t a DPS increase by itself, but it mitigates some of the DPS loss caused when you have to WoG yourself.  That’s still likely to be less effective than either of our top two glyphs though.  Final Wrath is a nice DPS boost during Execute, but it’s overall effect on a fight is much weaker.  Harsh Words is interesting, in that it’s a DPS increase whenever we’re not getting the benefit of Alabaster Shield.  Though it’s only about 400 DPS ahead without the T14 4-piece bonus (provided that even buffs the offensive capabilities of WoG – that’s only an assumption at this point), and since it’s only useful when we’re off-tanking it’s not as potent as our other glyphs in a practical situation.

The stat scaling simulation produces pretty plots, so I’ll post one:

At high Vengeance, hit and expertise are our top stats, followed by haste.  Strength, AP, and crit all trail the top 3 by pretty significant margins.  Of course, once you reach hit cap or spell hit cap, things change a bit.  If you stack 7.5% hit and 7.5% expertise, then the rest of your itemization should be focused on haste and strength if you’re looking to increase DPS.

Note that this advice changes somewhat in a low-Vengeance situation (off-tanking, 10-man raiding).  I won’t reproduce the plot, but in those situations strength and AP outperform haste and roughly equal the value of expertise.  Hit is still ahead by a bit, but not a lot.  So a hit>exp>strength strategy would be a DPS maximizer at low Vengeance.

There’s a weapon simulation as well, but there isn’t much to say about it.  Kilrak is very good (all versions), especially once you can put a gem in it.  There aren’t a lot of readily-available “tanking” weapons (str+stam+ one of dodge/parry/hit/exp/haste), though there’s a BoE available if you have the cash.  However, there are a number of “DPS” options with crit on them, and even a few agility weapons that have useful secondary stats.  But your best bet for weapons this time around is to get into normal-mode raiding.

Q&A Summary

I’ll try and post the answers to any significantly complicated questions here rather than fight with WordPress’ comment system.

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

15 Responses to MATLAB DPS Simulations for 5.0.5 and Q & A

  1. Jazende says:

    I’ve tweeted @Ghostcrawler but I haven’t received an answer, how would you feel about SS AP updating on shieldproc to remove the “need” to reapply it often at the start of a fight?

    • Theck says:

      I assume you mean Sacred Shield’s absorb value updating when the absorb bubble is granted rather than when the 30-second buff is applied? Yeah, I agree, I think that’s more logical given how Vengeance works. Standard HoT mechanics are that everything is set when the buff is cast, so SS is technically consistent with that. But most healers don’t have huge fluctuations in their spellpower multiple times within a HoT’s duration, either.

      It seems like it wouldn’t be hard to fix though. Since Sacred Shield generates a second, separate buff rather than a true “tick” of a HoT, the absorb calculation would just need to be moved from the primary buff to the secondary buff. I think.*

      *Note that I’m guessing a bit about how their code works.

  2. Loliinacan says:

    I always thought a tank’s primary job is survival>DPS(threat). Most of the choices you discuss seem to prioritize DPS>survival. I know the point of this post is discussion of DPS rotations and choices, but I am wondering where to draw the line.

    Specifically, you said: “I’d say the default choice is Execution Sentence, but I can see each of the talents being chosen for specific encounters.” It seem Holy Prism or Light’s Hammer would always be the better choice if our survivability is ever in question. I know that they are somewhat difficult to fit into our rotation, since sacrificing Holy Power is sacrificing survivability as well. Even so, I would like to be as survivable as possible. Is the idea that the healing usually be overhealing, or are they difficult to use reactively since they are on the global cool down?

    • Theck says:

      Well, we’ve always been balancing survivability and DPS. DPS checks are especially prevalent on heroic content, and we’ve been making choices based on that all expansion. For example, most of the early Ultraxion kills involved tanks wearing lots of DPS gear, because the fight didn’t stress their survivability too much to begin with.

      It becomes a very interesting question now that we do a significant amount of damage. In previous expansions, our DPS contribution was small enough that it was hard to find a large enough DPS increase to sacrifice *any* survivability. Now we have quite a few trades where both sides are significant. I don’t think the “as survivable as possible” paradigm is as important to progression raiding as many give it credit for. It’s important to be able to survive, but the best tanks aren’t the ones that ignore DPS entirely. They’re the ones that make sure they’re as survivable as they need to be for the boss while still contributing as much DPS as they can, by making the most efficient trades.

      Specifically talking about the L90 talents, though, I don’t think choosing Execution Sentence is a big swing either way. If you’re prioritizing any of the L90 talents below your Holy Power generators, then you’re not sacrificing any SotR uptime. So the survivability difference comes down to which L90 talent you choose and how you use it. You could argue that Holy Prism is a better survival tool since it gives you an ~100k instant-cast heal on a short cooldown. That’s not bad as a reactive tool, but we already have WoG for that purpose too. And we could use ES that way for a ~450k HoT if we needed it.

      Neither of them is a bad choice, but ES gives us a huge chunk of burst damage that can be helpful for DPS-critical phases, while Holy Prism is a weaker copy of a tool we already have (WoG). There might be fights where one is clearly better than another, but I don’t expect any of the L90 talents to be necessary on an average fight. As such, I’d take the one that gives me the most offensive benefit.

  3. Propal says:

    Have you done any Monte Carlo simulations on the whole “mastery package” damage smoothening vs the TDR benefits of the avoidance gearing (that some guides (like mmo-champ) seem to prefer, and if so, how big difference is there between the gearing method?

  4. Vayacondios says:

    Have you worked up (or know of anyone who has worked up) a lv 90 BiS pre-raid, reg, and heroic, gear list?

    • Theck says:

      I haven’t, and I don’t know if anyone has yet. You might want to check the maintankadin gear forum and see if someone has made one there.

  5. Pingback: The Light and How to Swing It: Leveling from 85-90 as a tank | DaPhat.com Online Gaming Community

  6. Pingback: The Light and How to Swing It: Leveling from 85-90 as a tank | Mikes Wow Macros

  7. Pingback: The Light and How to Swing It: Leveling from 85-90 as a tank - Owned Gamers

  8. Pingback: Avoidance, Mitigation, and Damage Smoothing | Sacred Duty

  9. Jamie Turner says:

    Theck i was curious if you had done any work on seeing Glyph of word of glories effectiveness in AoE situations I imagine that it would take quite a few mobs for it to be an increase over SotR but it plus one of our 2 lvl 90 AoE abilities might help us get some burst AoE threat for situations where we start tanking the mobs at low vengeance and the DPS needs to start AoEing them hard and fast

    • Theck says:

      Unfortunately, I didn’t get time to finish writing the AoE DPS simulation. I’m sure I’ll get to it, but probably not for another week or so.

      However, we can make an educated guess. In 6 seconds we need the extra 9% damage from Cons and HotR to equal one SotR. If we estimate 1 SotR = (2 HotR + 2/3 cons) on a single target (I don’t have numbers in front of me ATM, but this seems like a reasonable ballpark guess), then to get about 10% of the right hand side to eual the left, we’d need to be hitting 10 targets.

  10. Pingback: WoW Paladin Guide: Leveling from 85-90 as a tank | FREE World of Warcraft Guides

Leave a Reply