[4xShammy Arena] ToW Theorycrafting
I thought I'd open up a thread to do some theorycrafting (and real world discussion) about the choice between Totem of Wrath and Nature's Swiftness for a 4xElemental Shaman team. Here are (I think) the key tradeoffs:
Choosing Totem of Wrath over Nature's Swiftness:
+ Extra spell crit (to-hit seems largely unimportant) of 12%
- No nature's swiftness for Chain Lightning (or a heal)
- Totem of Wrath shares fire totem slot with the popular Fire Nova Totem (so its use once engaged with a melee might be limited)
+ Don't have to respec for Arenas if you like to PvE with ToW
The conventional wisdom is that Nature's Swiftness is too good to let go for Totem of Wrath, with PvP Resilience often being thrown out there as a counter to spell crit. This may be true, but I'd like to kick off the discussion with the strawman argument that ToW may be even more useful than people think due to resilience.
But first, some statistics!
I'm going to use Bernoulli Trials here to look at "binomial events". That is to say, when you cast a spell you either score a critical hit (with some probability, say P=0.20) or a normal hit (the other 80% of the hits, ignoring misses for a variety of reasons). When you play your 4 shammies and cast 4 lightning bolts, the important thing for PvP is how much burst damage you do when they land. What we really care about is the likelihood that we will have single, multiple, or NO crits out of that spell volley. Bernoulli trials can tell us some useful information about this unique question.
For example: Members of a 4xShammy team each have a crit percentage of 20%. If they all cast lightning bolt, there is...
a 59% chance that at least one of them crits. (Number of crits between 1 and 4)
a corresponding 41% chance that no one crits
an 18% chance that you get 2+ crits
and a very, very unlikely shot at everyone critting (about 1 in 625 volleys)
From a PvP perspective, again, these multi-crit chances are very interesting to us because they allow the target to take much more damage and possibly die in just 2 global cooldowns. So, let's try some interesting scenarios and compare them against each other.
Scenario 1:
Reasonably well geared PvP shaman group (20% crit) fighting other reasonably well geared opponents (400 resilience, say -10% crit chance)
OR
Not so uber geared shaman team (15% crit) fighting not so well geared opponents (200 resilience, -5% crit)
Net crit chance: 10%
Scenario 2:
Same as above, but with 4xToW added (+12% crit)
Net crit chance: 22%
Scenario 1 (no-ToW):
Chance of no-crits on 4xSpellcast: 66%
Chance of 1+ crit: 34%
Chance of 2+ crits: 5%
Scenario 2 (ToW):
Chance of no-crits on 4xSpellcast: 37%
Chance of 1+ crit: 63%
Chance of 2+ crits: 21%
To make it more interesting, extend the spell volley to 8 spells (say a LB+CL opening attack)...
Scenario 1: Chance of 3+ total crits: 3.8%
Scenario 2: Chance of 3+ total crits: 24.8%
You may be thinking/asking about Lightning Overload (the talent, not my blog ). Since LO has a 20% chance to proc, you can use the first example to think about what you'll expect to see for LO procs. 59% chance that at least 1 shammy gets a proc, 18% chance that two or more get a proc. I haven't done this yet, but it would be possible to build out an event tree to capture LO procs and their corresponding crit chances.
At any rate, with just a little bit of theorycrafting, it would seem that stacked ToWs would provide a nice increase to the chance that you'll get good burst damage on each spell salvo. The 2+crit events happen at a 1 in 5 rate as opposed to a 1 in 20 for the hypothetical examples above.
At the end of the day, I think the NS+EM+CL button offers many benefits over ToW, but it might not have as much of a commanding advantage over the ToW stacker as we once thought. For a player who is mixing PvE and PvP challenges, ToW may provide a good return on your investment and not require constantly respeccing. ToW may also be superior for BG players who are operating in a defensive or "totem fortress" configuration where they have long periods of time with many opponents (i.e. multiple salvos).
For fun, if you consider a very different case, say, Ellay's guys + 4xToW vs. "scrubs" in the BG with 150 resilience, his crit percentage would be about 29.25% (21+12-3.75). For an "Ellay vs. n00bs Salvo", that's:
Chance of no-crits on 4xSpellcast: 25% (47% w/o ToWs)
Chance of 1+ crit: 75% (53% w/o ToWs)
Chance of 2+ crits: 34% (14% w/o ToWs)
Chance of 3-4 crits: 8% (1.8% w/o ToWs)
(Bernoulli Trials tool at http://people.hofstra.edu/Stefan_Wan...bernoulli.html )