It's niche being SAM is lacking in DPS contribution compared to NIN, and any quarter decent group will be running a NIN thanks to TA. Sam is unwanted currently due to lacking raid utility and dps compared to its cohorts.
Printable View
I'm sorry, but this isn't true.
SAM brings enough damage to warrant a spot on his own.
The core of almost all groups, in respect to dps is:
NIN(always there)+RDM(apparently always there too)+BRD+SAM/MNK. On some very rare occasion, you'll see NIN+MNK+SAM+RDM.
The truth is that there's not a lot that can compete with SAM. The two other casters are both horrid in terms of personal dps and have poor synergy with the physical crowd.
That leaves DRG and MCH. These two... also have horrid personal dps, and inferior utility to NIN/BRD/RDM. They have more utility than SAM, but it's not nearly enough to close the dps gap.
That's why the fastest kills are pretty much "grab 2 out of SAM/MNK/BRD and add to NIN+RDM".
SAM is relatively desirable right now.
If you read back on the conversation, you'll see that the 'niche' I was referring to was the possibility that someone would pass up a NIN in favor of SAM, implying either a single melee set up or SAM/MNK, and the implications there of, hence why I was comparing NIN and SAM exclusively rather than all the melee. After all, even the most casual static that doesn't care about the meta are still gonna prefer a NIN for TA.
But shouldn't it be NIN - SAM a combo that should be inneficient compared to NIN - MNK/DRG or SAM - MNK/DRG, since slashing debuff doesn't stack NIN - SAM combo should be inneficient compared to the other ones.
The fact that NIN - SAM is one of the best combo only means that or NIN is broken, SAM is broken or both are broken.
There should more than one meta. 3 Casters should have a parallel meta, rivaling the physical meta.
But with so many physical dps buffs and only Rdm of the casters getting love in 4.0..,.. not possible.
Document: https://www.reddit.com/r/ffxiv/comme...ps_comparison/
Abstract: I did an analysis on the 80-percentile damage-to-boss statistics on Susano EX and Lakshmi EX from FFlogs to examine the raid dps contribution of 4 melee DPS. The analysis concludes that the raid dps contribution difference between DRG/MNK and SAM/NIN is only 4.5% of SAM personal dps and the actual difference in real fights should be smaller due to the amplification effect of stacking party buffs.
Note: Please read the sections regarding NIN, DRG, and MNK first, then refer to the Assumption and Preliminary Calculation sections for more details lest you are bored to death half way through.
Request: Please let me know if there are any egregious mistakes in the assumptions, methodology, and calculations. Any feedback is greatly appreciated.
I have issues with your analysis already xD
You basically make strong independence assumptions which probably don't hold.
You discount the slashing debuff utility of these jobs on the premise of "it's likely that the dps bring these if the tanks don't".
But these are not independent- it's because the strongest dps combination always has a slashing debuffer that you can afford to go PLD/DRK.
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).
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).
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 (Devotion? Something like that qq), single target balances (sometimes you just get these and you need to dump them on someone) all work much better on a SAM. In some circumstances, it may be that the aoe multiplicative effect of buffs favour a all-around utility comp, but in some others (especially with those I've named), they flavor having a powerful single target dpser.
If you want a statistical look at it, there's a very strong correlation between NIN+SAM cores and very fast clears according to FFlogs, going by the top 20/50/100.
Does this imply causation or is it an anomaly?
Imho, it's a case of causation (that core works better because it's inherently stronger).
Caster-statistician rant out.
(That said, I really appreciate the effort and the fact you neatly compiled this wealth of good information is very much appreciation. Keep it up, friend :) )
If the conversation is about a meta in a raid setting, and it is, then NIN is assumed to be in the party. Any group that cares about such things won't run without one. It has strong enmity manipulation, which is helpful for increasing Tank DPS through minimizing enmity combos needed. Somewhat relevant with healers, in relation to Smokebomb. TA is reliable, and it's DPS is at, or higher than a good portion of the other damage dealers.
At best, if we're talking PF/DF conditions (which doesn't matter for a meta conversation), SAM being the sole Slashing provider only benefits the two Slashing Jobs that don't have the ability or requirement to provide it for themselves- PLD and DRK. Assuming 2500 sustained for both Tanks over the course of a fight, that's ~500 rDPS contribution. Swap places for a NIN, and guess what? Still 500 contributed. The takeaway being- It's not exclusive to SAM, therefore it doesn't matter. Any argument saying that SAM gives a lot to a party with Slashing, can really be translated to the same argument with NIN. Except NIN has more than just slashing, so... yeah. The only imbalance here is that WAR is the clear loser, since it'll only ever, at most, benefit it's co-tank.
The real crime here is that NIN deals comparable damage to SAM while earning an extra 500 raid DPS that SAM has no answer for, using an extremely low balled estimate, through TA.