Results -9 to 0 of 3099

Dev. Posts

Threaded View

  1. #11
    Player
    Galvuu's Avatar
    Join Date
    Jun 2017
    Posts
    637
    Character
    Galveira Vorfeed
    World
    Ragnarok
    Main Class
    Pictomancer Lv 100
    Quote Originally Posted by SunnyHirose View Post
    snip
    You do have access to the spreadsheets with the data points (they were on the reddit thread), the most recent formulas and even the C# code for the simulator.
    Naturally, we must put "good faith" that the simulator is reasonably accurate and that there the functions that actual produce the data points (that we have no access to) are reasonably smooth (and even then, people did figure out the speed tiering).

    That said, on a fundamental level, I agree with you.
    I'm of the opinion that it's important to understand on what grounds we chose a secondary stat in detriment of another- people should understand that numbers like "stat weights" are linearizations around specific points, that you incur a greater error the farther you get from that point, that the quadratic and hyperbolic behaviour of stats may or may not favour stacking around certain values... things like that.
    People also need to understand the limitations of the simulator in capturing finer details inherent to some stats (speed, and how certain breakpoints that allow you to clip certain buffs would be invaluable- that next tier may not just be 0.01 gcds shorter, it may result in an extra Fire IV getting the effects of the Balance or Ley Lines) or even decision making (I don't know how it behaves regarding things like proc management; probably not so well, due to oddities in the 3.4 SCH weights).
    You should use information like this as a starting point, understand the formulas, grab Matlab or a sheet of paper or whatever, crunch some "napkin numbers" and hit a dummy for a couple of hours for experimental verification.
    I do this, I have tens of hours of test parses- in dummies and fights- and like 30 Ariyala gearsets. I like this kind of thing, and I fundamentally like to understand what's going on in general.

    But I also adequate the answer to the question x)
    Not everyone has the time, dedication or simply interest in that sorta thing. Spending hours to get a measly 1 or 2% in performance assuming you execute perfectly isn't for everyone. And that's fine, different strokes for different folks.
    So if I'm asked "what secondary yields the most damage?", the question inherently entails that the person asking isn't aware (or perhaps isn't interested) in these finer points. So I give a simple, approximate answer that should be better than randomly allotting the secondaries, but not as good as truly min/maxing them.
    Now, if some asks what's the best set in terms of secondaries that also meets a speed tier so that both Ley Lines and the arrow clip on at least the Kth spell, this is a harder question that warrants a deeper answer.
    Either way, I don't mind people who want simple guidelines to follow without much thought.

    Tl;dr: I mostly agree with you, but understand if someone wants a simple and effective answer that's neither completely min/maxed, nor goes in-depth into the inner workings of things (aka: if some wants a secondary stat heuristic black-box xD).
    (1)
    Last edited by Galvuu; 07-12-2017 at 10:48 AM.

Tags for this Thread