I have been putting together a spreadsheet to model fan of knives damage and sustained dps, and have decided to share my work with the hope of refining the model through wider scrutiny.


Goals of the spreadsheet:

• To help determine the optimal combination of weapon types, speeds, talents, and ability usage for varying AoE situations.
• To refine the fan of knives model for use in future analysis.

Notes on the model:

Currently the spreadsheet only models the dps from FoK (and poisons proc'd by FoK) alone, though I have done my best to account for multiple sources of energy regen.

I modeled the focused attacks procs from fan of knives itself by effectively reducing its energy cost, while focused attacks procs from autoattacking add to the energy regen per second field.

Slice and dice is present as a binary state to affect the energy regen from autoattacking; the model does not consider the cost of maintaining slice and dice, but some elements may be valuable in additional analysis of its maintenance cost.


Notes on the spreadsheet:

All the settings are controlled on the "Main" sheet. Most raid buffs/debuffs are assumed, but there is the option to disable sunder or faerie fire. Most settings are handled with a binary value (1 for true, 0 for false), but some talents use the number of spent talent points.

I obtained most of the stats used by entering my gear into Aldriana's Combat spreadsheet then manually entering the values into my spreadsheet; armor penetration and haste use paper doll numbers.

Some of the buffs are split into two options: the sustained increase granted by using that buff on cool down, or the effect of the buff on FoK dps while the buff is active.

The sheet will automatically determine whether to use wound or instant poison based on the damage per hit, but enabling deadly poison will force its use in the off hand.



I would appreciate any and all feedback or assistance verifying the accuracy of the formulas in the calcs sheet. I apologize in advance for the low ease of use, but I tried my best to make the formulas easy to understand without excessive abstraction into multiple cells.

Please feel free to let me know if you have any questions about the current model or how to use the sheet.

It is not a bad start; however, you should consider including the damage cap on AoE spells. For instance, in Onyxia 25 man, the start of phase 2 has a good number (>20?) whelps. When I head in to FoK the damage does cap out, but the poison damage does not. Also the energy return from the T8 bonus makes it possible to energy cap me every GCD.

I've updated the spreadsheet to include the 3.3 AoE cap mechanics:
The pre-crit damage per hit is modified as follows:
dmg*10*(1/Max(10,Target_Count))

Poison damage remains unchanged.

I've also placed a cap of 1 FoK per second to account for the gcd.

One thing I have been wondering is whether HnS can proc multiple times from a single FoK?

Onyx whelps or Yogg+0 would be suitable scenarios to test it, but alas I don't have proper swords.

To the best of my knowledge it can't.
This test was done over an hour using nothing but Fan of Knives while facing away from my current target while hitting 4 targets with Fan of Knives: World of Logs - Real Time Raid Analysis.
I didn't calculate the level reliatibility of the test but there isn't a single occurance of 2 successive Hack and Slash procs.

Wouldn't facing away from your target prevent HnS procs from hitting it? I'm not sure, but I would imagine HnS interaction with FoK would be similar to blade flurry's interaction, where FoK attacks would proc HnS on your target if it's in melee range.

I'd imagine the best way to test this would be to FoK multiple targets while in melee range without autoattacking your target, and use an axe in one hand, switching after multiple iterations to see if it can proc only from specific hands.

Edited due to more information recieved

And i di in fact like the idea this spreadsheet brings up. You do in fact have the ability to switch weapons in mid combat, so a rogue looking to fully optimize himself (lets say in Onyxia) can use his/her usual weapon combo for the first phase, then when whelps spawn, pop a macro switching weapons that are already poisoned before the fight.

Excellent post sir.

You'd expect 8 FOK hits to produce a HnS proc 8*0.05=40% of the time when using FOK and a double HnS proc 8*0.05*0.05=2% of the time. With 500 succesfull FOKs in a row the chance of not seeing a single double proc is 0.98^500=4.1*10^-5 or 0.0041% while expecting to see 10 double procs in total.

Now does this mean we can never see 2 HnS in a single FoK? or is there a chance one mainahand hit can proc one and one offhand hit can proc one?

So, if I'm understanding Bural's analysis correctly, then it follows that the expected outcomes when performing FoK with double HnS weapons is:

* Gaining Double HnS : Zero (chance should be 0.05*0.05, but only 1 attack happens)
* Gaining No HnS: 90.25% (0.95*0.95)
* Gaining Single HnS: 9.75% (1 - 0.95*0.95)


Edit:

Upon some testing I did, I'm suspecting (because I didn't track the exact count of attacks) that HnS can only proc from the main hand FoK hit. Summary of the test:

- Stood behind target dummy, axe equipped in mainhand, a dagger in offhand
- In melee range of first dummy, but not auto-attacking, in FoK range of one other dummy, and just spamming FoK as energy allows
- Didn't take long before I saw "gain Hack and Slash" in combat log

- Repeated the experiment with the axe in offhand, dagger in mainhand
- Could not produce a HnS in the combat log for a good 5-6min of testing. Average ~4sec between FoK's that's ~15 casts per minute x 2 targets hit. So roughly 5x15x2 = 150 FoK attacks with the offhand axe
- Chance for HnS to not have happened in 150 attacks is .95^150 = 0.000456

So if true, the outcomes of a FoK with axes would simply be:
* Gaining HnS: 5%
* No HnS: 95%

This also suggests that if we're weapon swapping to a slow offhand before FoK-heavy phases, then *any* weapon type in the offhand would do about just as well.

While performing some personal analysis on the usefulness of [Heartpierce], it occurred to me that energy regeneration likely scales better in multi-target situations than it does in single-target situations, so I decided to investigate.

In my spreadsheet, I started with a set of baseline stats similar to my own, then added the stats that Rib Spreader has in excess of those on Heartpierce.

That yielded roughly a 250 dps increase on 5 targets over the baseline stats.

From the baseline, I manually increased the energy regen by varying intervals until the dps matched that of the rib spreader stat increase.

The break-even point was roughly a 38% up-time for the buff.

I have not done personal testing, but I would imagine if each FoK hit has a chance to proc the buff, that up-time would easily exceed 38% with sustained use of FoK.

This leads me to believe that Heartpierce is likely the best weapon to use for FoK. It would most likely be best used with a weapon-swap macro to place in the off hand with instant poison before an AoE pull, but there may very well be a risk of inefficient proc use if the duration of the pull isn't sufficiently long.