Elemental, Shaman, Theorycraft, WoW

Elemental 3.3.2: Buffs and Bizuri’s

With the translated Korean patch notes for 3.3.2 out, it seems reasonably certain that we know what’s coming for us elemental shamans.  Let’s take a look at the changes to our tier 10 four-piece bonus and our shamanism talent!  I also want to comment very briefly on Bizuri’s Totem of Shattered Ice.

Tier 10 4-piece

This has gone through a few conceptual changes now, but it looks like it’s finally been decided.  Though we’ll have to wait for English patch notes for precise wording, the new bonus is:

Your successful Lava Burst casts increase the duration of your Flame Shock on the target by 6 seconds.

If this sounds familiar, it’s because the same mechanic is already used for many similar effects in-game including the interaction between a Shadow Priest’s Mind Flay and Shadow Word: Pain spells, or an Affliction Warlock’s Haunt and Corruption spells.

It’s also very similar to what our two-piece tier 9 set bonus already does – increasing the duration of Flame Shock (that reminds me, I’ll need to update that post again).

The benefit this gives us is that we have to cast Flame Shock less often.  Because Flame Shock does a lot of damage over time but very little damage up-front, this means that we can cast more higher DPS spells instead of Flame Shocks while still benefiting from the spell’s extremely high damage-per-cast-time.  You can think of it as saving the effective cast time of a Flame Shock every 18 seconds or so, while also benefiting from having the DoT still tick uninterrupted.

In a stand-up fight

Is this gonna be a standup fight, sir, or another bughunt?

For a stand-‘n-nuke scenario like Patchwerk or Festergut, the bonus is actually pretty big.  Modelled in the spreadsheet, Flame Shock can tick 13 to 15 times – which is a duration of 39 to 45 seconds, and potentially more at certain haste values.  The range of damage increase depends again on gear level, but for most people it’d be somewhere between 150 and 250 DPS which compares very well to our previous set bonuses.

In a bughunt

Fights with high levels of movement are a bit of a different picture.  Normally while moving one of the first things we’d do is refresh Flame Shock.  However if we do this with 4t10 we could actually lose DPS – a Flame Shock cast immediately after a Lava Burst can end up expiring sooner than the Flame Shock which was already on the target and being extended!

However, we have another shock option while moving: Frost Shock!  While it does cause somewhat more threat than its fiery cousin, and it only has a 25 yard range, it also does a lot more up-front damage than Flame Shock and doesn’t reset the DoT timer – so not only are we getting more up-front damage per hit, we’re also getting an extra tick of the DoT.  If we assume that we’ll be able to get about 40 seconds out of every Flame Shock and use Frost Shock instead during movement phases, the bonus is potentially worth about 100 DPS even on a fight with lots of movement.  Of course if we don’t use Frost Shock and are moving more than every 20 seconds or so, we get no benefit at all from the bonus.

Where the bonus unfortunately does look bad is on multi-target fights where targets are unlikely to live long enough for the DoT extension to be any use.  And even if there’s one target alive the whole time, chances are our Lava Bursts will be needed on the adds and thus the DoT on the main target won’t be extended either.  Lady Deathwhisper is the obvious example here, and perhaps the Blood Princes too – though I reckon the latters’ empowered phases last more than long enough to get some serious mileage out of the bonus.

Overall it’s much less ambiguous than the old Lava Burst cooldown reduction and makes 4t10 very desirable over alternative options.  Unfortunately, however, it does look like we’ll still have to “clip” the last few seconds of the DoT to avoid it falling off during Lava Burst casts 😦


Since 3.0.8 when it was added, this talent has been the core of our high-end scaling and was tweaked up in 3.2.2 when our damage was flagging a bit.  It seems that in 3.3.2 it’ll get another 1% per talent point added to its existing bonuses.  This means that with 5 points the bonuses are as follows:

Spell Current benefit 3.3.2 benefit
Lightning Bolt 15% (3% per point) 20% (4% per point)
Chain Lightning 15% (3% per point) 20% (4% per point)
Lava Burst 20% (4% per point) 25% (5% per point)

Shamanism affects the co-efficients of our spells, which is how much of our spellpower is added to their base damage when we cast.  What this means is that an additional 5% of your spellpower will be added to every cast of Lightning Bolt, Chain Lightning or Lava Burst.

Spell Current co-efficent 3.3.2 co-efficient
Lightning Bolt 86.43% 91.43%
Chain Lightning 72.14% 77.14%
Lava Burst 77.14% 82.14%

Of course Lightning spells also Overload, and despite the base Overload spell only doing half damage, they do gain the full benefit from shamanism:

Spell Current co-efficent 3.3.2 co-efficient
Lightning Bolt Overload
50.71% 55.71%
Chain Lightning Overload
43.57% 48.57%

Overloads fire at a 33% rate for Lightning Bolts and an 11% rate (per target hit) for Chain Lightning, so the total benefit of our spellpower now going to each (non-crit) spell is:

Spell Current total SP 3.3.2 total SP
LB+Overload 103.16% 109.81%
CL+Overload 76.93% 82.48%
Lava Burst 77.14% 82.14%

Of course the base damage of each spell remains unchanged. Including overloads the base damage figures are 896 for Lightning Bolt, 1099 for Chain Lightning and 1355 for Lava Burst.  This means that the effect of this change will be felt most at higher gear levels where more of a spell’s total damage comes from its spellpower.  For example presently, at 2,000 spellpower, only about 70% of Lightning Bolt‘s damage comes from SP:

103.16%*2000 / ( 896 + 103.16%*2,000 ) = 69.7%

But at 4,000 spellpower, it’s over 80%:

103.16%*4000 / ( 896 + 103.16%*4,000 ) = 82.2%

I won’t bore you with any more maths here, but at tier 9/tier 10 gear levels, the spreadsheet is putting the tweak at between 4-5% extra overall DPS in all situations, with 5% being the absolute maximum in Best-in-Slot gear without using fire DPS totems.  As the only spell we regularly use that isn’t affected by this change is Flame Shock, it’s a very reliable boost to our damage in all situations and doesn’t affect enhancement or restoration.  It’s a very safe “knob” for the devs to turn up a notch.

Bizuri’s Totem of Shattered Ice

This 30 Frost Emblem relic is a clear upgrade over its Triumph Emblem sister.  But how much of an upgrade?  Obviously Flame Shock will be ticking on something for way more of a fight than we’ll be casting Lightning Bolt for, and the buff lasts more than twice as long (30s vs 12s), so it’s more reliable.  But then it also has that ramp-up time at the start when the Flame Shock DoT has only just started to tick which we simply can’t circumvent.

Agash did some maths suggesting that the average uptime of Totem of Electrifying Wind is (roughly) between 90 and 99% with about 95% uptime during movement.  If we generously assume an average 97% uptime, that works out as 194 haste.

While Bizuri’s buff won’t ever fall off (30 seconds duration and stacks on any Flame Shock tick), it does have a fixed ramp-up time of 15 seconds.  In a 5-minute fight that means the uptime of the full 220 haste is only 95%, but the uptime of at least a single stack of the buff is 99%.  This also works out as an average 97% uptime for the full haste:

( (3*0+3*44+3*88+3*132+3*176)/15 + 220*(300-15) ) / 300

That makes the average haste granted by the totem about 213.  Of course it’ll be less in a shorter fight because the 15-second ramp-up time remains constant.  Actually, in theory, a fight shorter than about a minute might favour Electrifying Wind due to the possibility of higher total uptime.

But for the 5-minute fight, if we assume a worst-case scenario for Electrifying Wind and say it’s only got 90% uptime (thus an average 180 haste), then the benefit from equipping Bizuri’s Totem of Shattered Ice is 213 minus 180 equals 33 haste, or 1.1 haste rating per Emblem of Frost spent.  Another way of looking at it is about 1 DPS per emblem.  Thus in my opinion it’s still better to save it until you’ve got tier 10 (especially with the 4-piece change), even if you only have access to the ilevel 251 stuff (as then you’ll probably be wearing ilevel 232 stuff to upgrade).

And no, I’ve not analysed the per-frost-emblem benefit of going from tier 9 to tier 10, but if anyone wants to try I’d be most interested 🙂



15 thoughts on “Elemental 3.3.2: Buffs and Bizuri’s

  1. I think the dps per emblem of each item is going to vary a lot, if you take into account the differences in current gear for individual shaman. For example those of us who cannot shake the sodding Deathchill cloak will probably get the most mileage out of one of the emblem cloaks (sadface) than from upgrading their triumph relic to the frost relic :/

    Posted by pewter | January 28, 2010, 4:40 pm
  2. Well done, informative post.

    I was guessing a 4% DPS upgrade from the bump in Shamanism, but it’s great to see a walk-through on the math.


    Posted by Vydra | January 28, 2010, 4:53 pm
  3. Please forgive my basic question, I follow everything else, but I do not understand how, and I’ve seen it mentioned on EJ as well, we *lose* DPS when refreshing FlS during movement.

    If there are <17s remaining on FlS (my LvB is almost 1s) shouldn't resetting it specifically do *no harm* (ie you’d be the same off as if no T10 bonus)? I understand of course, you would not get a *benefit*, but actually *lose* DPS? The only way my brain is letting me look at it is, as long as we cast FlS as little as possible while standing still, we are getting a benefit, refreshing it when moving is/was just a necessity to do *something*, but, how is this different than before? Unless…something to do with how long you’re going to stand; if you are running and know you have like, 10s on FlS, and that would last you another…4 LvB, before which you’d move again anyways, is that why you don’t need to refresh it? What if you have less than 10s, what if you have 5s, surely at some point it’s harmful to *not* refresh it when moving? Would some kind reader please explain it so I can understand?

    On a slightly side note, I practiced moving this week on Prof and using FrS in anticipation of the change, and it wasn't too bad at all! *So* many fights in ICC involve standing as far away from other people's cooties as possible, so it's pretty easy to just claim a patch within 25y without a problem, benefiting searing as well. Being close when poping the big guy is a nice bonus too, and of course, it allows you to use nova without running in some crazy distance only to have to drop 1 fire totem and run all the way back out.

    Posted by Acariel | January 28, 2010, 6:42 pm
    • The specific example I had in mind is using the BiS stats. Normally you’d have to refresh Flame Shock 40.38 seconds after the first cast. But if you cast a Flame Shock immediately after the second Lava Burst, you now have to refresh Flame Shock at 40.07 seconds (0.31 seconds earlier). So you have lost a tick of Flame Shock’s DoT and caused your next FlS to have to be cast slightly sooner.

      The reason I said “could be” and not “will be” is because the precise circumstances in which it’d be a DPS loss are very very specific, but the fact remains that there is a possibility – and that emphasises how important it is, with 4t10, to swap to Frost Shock while moving.

      Of course the reason we cast FlS on the move is because we can cast it on the move without sacrificing anything, whereas if we cast it while standing still we are “sacrificing” a cast-time spell. Thus if we’re forced to move while there’s roughly the same or less time remaining on Flame Shock’s DoT than there is on Lava Burst’s cooldown plus its cast time, it’s still better to cast Flame Shock as it’s now extending the timer and potentially saving a GCD while we’re standing still 🙂

      Some examples assuming a 1s Lava Burst cast time:

      Example 1: You are moving to avoid a slime explosion and your FlS DoT is at 7 seconds. You have just cast Lava Burst and its cooldown is now at 7.5 seconds. 7.5+1 = 8.5 > 7 so it’s better to cast Flame Shock.

      Example 2: You are running to a player with a spore. Your FlS DoT is at 16 seconds and your Lava Burst comes off cooldown in 2 seconds. 2+1 = 3 < 16 so it's better to Frost Shock (or even Fire Nova) if you can.

      Unfortunately it seems that with 4t10 we will nearly always have to recast Flame Shock while there's still a tick left on its DoT, so the trick will be deciding exactly when it's best to do this! I think a decent rule of thumb would be “refresh FlS if it’s at 10 seconds or less”.

      Posted by Charles | January 28, 2010, 7:55 pm
      • Thank goodness this is not EJ, and I am allowed to thank you. It never occurred to me to take into account LvB’s cooldown…and instead I was only thinking of it in terms of the theoretical max you can cast it (is if LvB was magically always available to be used when you stopped moving). Duh! Thank you so much.

        Posted by Acariel | January 28, 2010, 8:23 pm
  4. Told you they weren’t aware of or concerned about the Flame Shock clipping issue, Zamir!

    Posted by Bergg | January 29, 2010, 2:07 am
    • 😦

      Well I think the problem with the original bonus is that the clipping issue significantly reduced its DPS. I can’t think of any reason why the subsequent change would have been made other than that.

      This new bonus, though it will require some clipping, is almost unaffected by it in DPS terms. In fact clipping will guarantee we get the longest possible Flame Shock each time and thus may well be an aggregate DPS increase.

      (I do feel in general that it’s too easy to be critical of Blizzard and assume they’re idiots, so when a change like the one previously announced comes along, it seems appropriate to give them credit for it)

      Posted by Charles | January 29, 2010, 1:32 pm
      • Despite my many many criticisms, I’m looking forward to the current incarnation of the T10 4-set bonus. I think Elementals will be pleasantly surprised; I have a feeling it’s going to be a bigger DPS increase than we thought.

        Posted by Bergg | January 30, 2010, 11:12 am
  5. I’m a bit dumb in that I went ahead and bought the Bizuri totem as my first EoF purchase. I regret it a little, but I’m not losing sleep over it.

    Some pluses:

    – Better for pvp. Yeah, more often than not, FS will get removed from a target, making it harder to ramp up the haste, but when you do get the buff fully stacked, it’s fantastic for pvp. You don’t get to cast LB very often, so keeping the haste buff up from the 245 totem would be really difficult.

    – Enhance: Being that I’ve had just about an entire set of ICC-caliber enhance gear fall in my lap, I thought I would need to buy a badge totem for that set, only to find out that enhance shaman do use FS, and that haste is an amazing stat for them, too.

    No, these are not convincing arguments for someone that is purely a PVE elemental shaman, but it is food for thought. 🙂

    Posted by Kazgrel | January 29, 2010, 4:43 pm
  6. Thanks for this!

    You write a really good article mate 🙂 I’m glad the very small elemental community has you!

    Posted by pewter | February 1, 2010, 9:21 am
  7. hmmmm heres what i did to verify your math i took my draenei shaman into if and had 6 goes at the dummies. the totem of electrifying wind had an uptime of 75 %. i crossreferenced my procodile results with other blogs and found that even you stated an uptime of 70%, give or take, here:


    so either it is just a minor error or your math is wrong =)

    btw. if you pull fast enough or kill rats etc on the way u can keep the proc alive 4 killing trash xD

    Posted by hassan | February 2, 2010, 11:54 am
    • The proc chance is 70%, which is the chance that any given LB will cause the effect! The uptime is a function of proc chance, effect duration and number of LBs cast.

      Having said that, I didn’t actually do any uptime math for ToEW, I just used Agash’s findings 🙂

      Posted by Charles | February 2, 2010, 12:19 pm
  8. This is a bit unrelated to anything mentioned so far, but early testing on shadowpriest.com places the Phylactery of the Nameless Lich (formerly our best trinket) with a 90 second internal cooldown. I plugged this into the new version of ZAP! and changed it’s trinket cooldown to 100 (assuming 90 second internal, 10 second proc window), and it’s roughly in line with the Spyglass and Reign. In fact, the order went something like:

    Phylactery (Heroic) > Spyglass (Heroic) > Reign (Heroic) > Phylactery (Normal) > Spyglass (Normal) > Reign (Normal).

    So assuming the 90 second internal cooldown is correct, our BiS trinket is clearly the DFO (either version), then followed by three trinkets that are very close in relative value, which isn’t necessarily a bad thing, as it actually gives options for one of those slots.

    One thing to note, though, is that I haven’t done extensive testing with gear setups and the spreadsheet yet – I just looked at their individual relative values, so they may be subject to change.

    Posted by Shkarn | February 3, 2010, 11:47 pm


  1. Pingback: 3.3.2: It’s About Time « Bloodlust - February 3, 2010

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

Get emailed about new posts!

Join 27 other followers

We write about

%d bloggers like this: