Avoid quote splitting like this - it usually starts flame wars.
Originally Posted by Avitus
I'm not sure you're getting my point. When I say there are factors (lag, human error, fight mechanics such as movement) that debunk some very specific "blind" theorycraft (read: unrealistic), then I'm not at the same time saying that the sum of all theorycraft is useless and suffers from the same problem, you're mistaken here.
If, using your system under optimal conditions ("Blind theorycrafting") yields fewer casts over a given period, than using your system in actual play, given the lag, human error, fight mechanics, etc, must necessarily result in fewer casts over a given period. If you disagree, please explain how it could possibly be otherwise.
A constant delay on CS, every single cast is a "predictable" penalty. It's not wild theorycrafting. It is quantifiable, if you go and hit something for 1 minute using J->CS (with 4-set) you'll know exactly what I mean.
It's an illusory penalty. That's precisely my point. I'm a big hands-on kinda guy, so I'll take you up on that challenge. (I did this some time ago, but I just went back to the dummy and did it again.)
After Several 1 min tests:
J-CS-Cons-DS = 8 J, 8 CS, 5 DS.
J-DS-CS-Cons = 8 J, 8 CS, 5 DS (But fewer ticks of cons).
There is no big win in the long run as you suggested earlier. Sticking a random ability in between J and CS basically means CS gets delayed for a couple globals every-other judgment cycle instead of one second each cycle.
Using CS-J-Cons-DS, on the other hand, results in 8 J, 10 CS, 5 DS, and the same ticks of cons as with the standard priority system.
However, building a magical cycle that fits perfectly on paper to guarantee some sort of reduced clash bonus falls apart the second any "realistic" conditions come into play (again: lag, human error, fight mechanics).
What magical cycle are you talking about? All I did was map out the typical FCFS system, one with your suggestion, and one using CS first. How is 'use CS before J' any more 'magical' than your suggestion of 'Use DS between J and CS?'
Again, map out a cycle like that for 6 minutes and I'll show you that it's pointlessly complicated compared to FCFS and provides identical results (only if it works out perfectly, in all other cases it's a loss).
Now what's easier to maintain: FCFS or a 21.5 second long rotation with arbitrary waiting in between? Especially since it provides no benefit over FCFS in return.
I guess you're referring to Banka's rotation, which had been dismissed before your first post on the subject. But maping out 6 minutes worth of a cycle that repeats every 21 seconds seems rather pointless in any event...