Quote Originally Posted by Galvuu View Post
This is already an issue itself, but I'm not comfortable just discarding this piece of utility because we implicitly assume that "it's always there" (which shouldn't be the case, but that's how centralising NIN really is these days).
This is a legitimate criticism given that PLD/DRK accounts for 42.3% of tank compositions. If the composition doesn't have WAR, then SAM/NIN has much higher raid dps contribution than MNK/DRG. Otherwise, SAM/NIN is only about 4.5% higher than MNK/DRG.

Quote Originally Posted by Galvuu View Post
Also 5% is a lot. If you raid, you've seen many low% enrages. Now think about that one week where you were stuck on A8S at sub 10% and consider that half of those would've been clears if you were using the meta comp.
You can't tell me to disregard this (and, seeing the slashing debuff comment, this is a slight underestimate under your questionable premise xD).
I think you mistook 5% of SAM dps for 5% of raid dps. On average, SAM does about 18.6% of raid dps. If MNK does 4.5% less than SAM (according to my analysis), i.e. using MNK in place of SAM would cause you to lose only 0.837% of raid dps. This is negligible.

Quote Originally Posted by Galvuu View Post
Another issue I have is that the multiplicative argument goes both ways- if you have a very strong base dps, you get more out of some buffs. Dragon Sight, Ifrit's single target thing...
Unfortunately, this also means that the dps data for SAM from FFlogs is inflated by single-targeted buffs concentrating on SAM. Since the data is aggregated, I can't really filter out the effect of single-targeted buffs; therefore, I cannot factor your concern in my analysis.