Twins Heroic and Blade Flurry:

Assume you have dark attunement, are attacking Fjola Lightbane. Your damage will be increase to 150% of your normal damage, if you were to switch to Eydis Darkbane on Shield of Darkness, your damage would dip to 50% (this may be slightly less ~47-48%) of normal. Staying on Fjola Lightbane and using Blade Flurry will cause 75% (1.5 * .5) of your direct, non poison damage to Eydis Darkbane. (again this might be slightly less ~70.5-72%)

Even in the case that only 70.5% of weapon damage lands on the BF target, it will yield a net damage increase, as long as poison/bleed damage is less than 29% of total damage caused, which should be the case for a evis only spec.

However, most other dps'ers have switched color and target, leaving you as the only darkie among a lot of whites, making you extremely vulnerable to incoming opposite colored orbs.
I assume "Goal-driven DPS Optimization" still means staying alive, even if you do less damage for 15secs.

You shouldn't be switching color, especially as melee, to DPS the opposite twin for a shield, unless it's the 2nd/4th shield and you've done so preemptively well ahead of time, so that isn't the issue. That said, are we sure that the Blade Flurry damage replication works like this? I asked earlier in the thread and didn't get an answer as to what order the multipliers are applied - if the copied damage is the Light/Dark attunement damage multipliers are applied after the Blade Flurry copy, then not switching would absolutely be a DPS loss on the shielding mob because you lose poison damage for no actual gain, aside from maybe not needing to reset your cycle (which is trivial with proper planning good play).

Add in Bloodlust, stars aligning and you get:

WoW Meter Online - Combatlog Replay

Ignore our other retarded attempts, just in light of the latest discussion in this thread I thought this was an amusing moment to show.

Noone changes attunement on shields in our strategy, melee does more damage by just switching target.


@feist-mok: i checked combat logs and my theory seems to be within a reasonable threshold of actual mechanics

Despite the fact that the portals are exactly 2min apart one must keep in mind that using KS requires not only positioning to have the KS hit the one portal and not portal+boss, but that seconds will be spent in doing this.

The time between portal spawn and add appearance is such that imho being glyphed for KS does help in this fight so far as it ensures that all rogues are in fact ready to KS single-target the portal when it pops, not when it is too late and the add is outside with it.

This nullifies the discussion as to whether BF is merited or not: if the CD on KS is ready and the rogue is in place the portal will be down in seconds as suggested. Whether or not the merits of this apply to a 25man with multiple rogues that could otherwise be glyphed for TOTT is debatable, but for 10man I'd say that KS would probably be a good way to go.

The idea of KS there is not to hit Jaraxxus at all, but getting the portal down asap, i.e. all KS hits should go on the portal only.