FD was last updated on 8/26 and is working off an earlier build. SimC was simming about 10k dps higher than FD for 17331 (the build that FD is using, I believe). SimC is now updated to build 17345, and Lokrick claims it is now more in line with BM mechanics on the PTR.
Also, SimC has always been more accurate at modeling haste than FD. Zeherah has certainly acknowledged that. I know that for both simulators, they are not accurately modeling MM rotations (Lokrick detailed his attempts at addressing this here), but I'm not entirely sure about the limitations of the BM profiles. We'll certainly see updates to the simulators in the next 3-4 weeks as Lokrick and Zeherah have more time to adjust to the 5.4 changes.
Edit: P.S. I am still recommending BS over AMoC on the BM guide. Such a small gain is negligible when considering in-game mechanics, the focus requirements of AMoC (especially given the increased focus cost of AS), etc. Besides, simulators tend to award more hits from pet-like abilities like crows and DB than what actually occur in average parses. We could be seeing something like that in SimC right now for AMoC.
im not a big fan of 30 sec dots that may not get wasted may get wasted. specially on adds fights. they really need to change those 2 talents AMoC and Lynx Rush. im still not sold on TOTH because the 3 gcd requirements i think is too much and it lock you down way too much, if they change that talent to reduce focus cost by 20 of next 3 abilities that cost focus would be much better and i will def take it over fevor or even dire beast.
As for Haste being hard to simulate for FD, it should not be affecting BS nor AMoC in any major ways. I concede that the overall DPS might be slightly off from the "real" value, but once again - with no haste procs etc, the amount of damage added by haste is fairly consistent apart from how many cobra shots etc we do, with no real or extremely minor effects on FD/AMoC.
But if there is any logic/math behind why AMoC would suddenly overtake Blink Strikes with 3K dps, I'm all ears tbh. I just can't for the life of me figure out why.
FD's haste system is based off the old plateau system. You can game the FD simulator to induce extra auto attacks, RPPM procs, and pet hits by adjusting the haste up or down by small increments, even though this is not how attacks work in game. Most people call this "shot shifting", and when I post FD BiS profiles on EJ, I tend to try to find these plateau marks, even though they're artificial. You can read more about testing of this mechanic in the DB section of this guide. This was also addressed in an update to Rogerbrown's guide on the Method site here.
Unless the FD engine is fundamentally changed, you will always see these plateau marks with haste, and that's why the stat scaling that is generated for haste sometimes wildly varies when you change haste reforges in your profile in FD. The nice thing about FD's model is that you won't get deviations when running the same profile. However, this simplification of haste does insert artificial plateaus in FD for many types of procs and attacks, even though the plateaus do not exist in the game for hunters (and this will affect all abilities, especially where RPPM trinkets are concerned).
SimC simulates thousands of runs without artificial haste plateaus. It tries to emulate the way the in game system actually works for hunters by having a random percent chance for each attack to generate an extra attack or proc. That's why the haste model is so much more realistic (in regards to in game performance) in SimC than FD. It's not perfect, and the modules are in a constant state of flux as developers try to approximate the in game engine, but fundamentally, it is more accurate than FD.
This has been a known problem in FD since the haste plateaus were removed from some hunter mechanics and was compounded by RPPM (which will affect both BS and AMoC valuing in FD). FD gives a good overall picture, but SimC will continue to be more accurate for RPPM trinkets (Zeherah's comments on this), pet attacks, and anything else that is not based on haste plateaus. If you're interested in seeing the kind of simplifications that occur in FD, you can check the DB calculations in Zeherah's haste calculator. The fact that FD and SimC are ever close is remarkable and shows how much work Zeherah puts into the tool.
Maybe because aMoC (.206) scales better with RAP than BS (.168)?
Unlike aMoC, BS scales with Haste though but new gear will improve RAP much more than Haste will affect WH uptime.
Not saying aMoC is truly better but it can make sense.
BS scales directly with every stat we have because it's pet damage. AMoC doesn't scale with haste at all AFAIK, which would imply that even if it had stronger AP scaling, it shouldn't be able to overtake.
Can someone bribe lokrick to fine tune the SV profile on SimC.. This is anecdotal but i did multiple 10 min tests with all buffs and lust, hast > crit for BM and crit > haste for SV and I was consistently doing anywhere from 10-15k more dps as SV.. CD reduc and ebon.. Which ofcourse doesn't mean much but it has me curious to say the least.. Even if they are equal, in real raid settings, SV should do slightly better cuz of target switching and whatnot.. Although i'm not a fan of the 4 pc SV.. my fingers were in legit pain after a few tests.. the most button spamming rotation I have ever encountered..