Page 1 of 3 1 2 3 LastLast
Results 1 to 10 of 22
  1. #1
    Player
    jrbaetz's Avatar
    Join Date
    Sep 2018
    Posts
    11
    Character
    Omlette Prime
    World
    Exodus
    Main Class
    Ninja Lv 90

    Crafting mechanics and probability engine...

    Crafting mechanics are screwed, as far I have counted the probability engine doesn't match up to the %'s displayed.
    (0)

  2. #2
    Player
    Shadygrove's Avatar
    Join Date
    Mar 2011
    Posts
    1,408
    Character
    Alya Mizar
    World
    Sargatanas
    Main Class
    Marauder Lv 72
    Its streaky. I sometimes run through 3 - 4 crafts with the exact same patterns of goods or fails.
    (0)

  3. #3
    Player
    MaybeOliverB's Avatar
    Join Date
    Nov 2017
    Posts
    161
    Character
    M'naago Cat
    World
    Gilgamesh
    Main Class
    Dragoon Lv 70
    Yeah, its because you're HQing a non HQable item.
    (0)

  4. #4
    Player
    jrbaetz's Avatar
    Join Date
    Sep 2018
    Posts
    11
    Character
    Omlette Prime
    World
    Exodus
    Main Class
    Ninja Lv 90
    I think if its going to say 80% it should be 80%. If gear plays a factor in chance to land something their should be a stat or some description in it. I use the lazy mans syneth, hasty touch steady hand method, that being said there is no reason to bust over half my synething process at 80%, thats over 50/50 which makes no sense by probability standards.
    (0)

  5. #5
    Player
    Frizze's Avatar
    Join Date
    Feb 2016
    Posts
    2,837
    Character
    Frizze Steeleblaze
    World
    Lamia
    Main Class
    Black Mage Lv 90
    Quote Originally Posted by jrbaetz View Post
    I think if its going to say 80% it should be 80%. If gear plays a factor in chance to land something their should be a stat or some description in it. I use the lazy mans syneth, hasty touch steady hand method, that being said there is no reason to bust over half my synething process at 80%, thats over 50/50 which makes no sense by probability standards.
    So this is an "issue" that has come up before. Someone notices that some of their actions are failing more then they think they should, so they make a post about how the RNG is broken and on and on. Is it possible there is an issue with the RNG model? Of course. That doesnt mean i believe there is one though. Players tend to jump to conclusions based on small sample sizes, and its human nature to remember all the times that things went against us more then when they went for us(i forget that i havent failed a 90% chance in 6 or 7 weeks, but ill tell everyone wholl listen when one fails three times in a row). The RNG of the game shouldnt be judged based on the last 10 items you synthed, or even your result over the last 10 days. To determine if the RNG is working as intended, you should be looking at the last 10 million actions and see how closely they match the stated percentage. If the devs want to run a report on this and let us know that everything is/isnt working correctly id welcome that. Failing that(i dont think its likely theyll share any such report), id tell you that every time a player has been convinced of this and done a focused trial theyve eventually reported that the percentages are more or less accurate(or theyve never shared any findings). If you record your next couple thousand results and do the math, youll likely see that everything is more or less working at the intended percentages.
    (0)

  6. #6
    Player
    MistakeNot's Avatar
    Join Date
    Sep 2015
    Posts
    2,245
    Character
    Auriana Redsteele
    World
    Zodiark
    Main Class
    Paladin Lv 82
    Quote Originally Posted by jrbaetz View Post
    Crafting mechanics are screwed, as far I have counted the probability engine doesn't match up to the %'s displayed.
    This is based on how many samples?
    Unless you have several hundred tries (as an absolute minimum) to compare against, you have no idea if the RNG is good or not.
    You would need many thousands of samples to be able to tell for certain if it is working correctly.
    (0)

  7. #7
    Player
    Alien_Gamer's Avatar
    Join Date
    Nov 2015
    Posts
    902
    Character
    Cynehild Westknight
    World
    Jenova
    Main Class
    Warrior Lv 90
    To put sample sizes into perspective, you've looked at a couple hundred maybe?

    When devs do a test sample they pull datasets in the 500,000 to 1,000,000 range. The RNG engine that crafting uses is the same one as everything else uses, stop and think about how many random elements there are in the game and multiply that across what everyone is doing in the world and DC. Think about how many buttons are clicked in those few seconds and you'll realize this isn't 10 successive fails in a row, this is 10 fails out of tens of thousands in the last 30 seconds and you just got unlucky to get bad results while others may not have.

    RNG is RNG.
    (1)

  8. #8
    Player
    Elamys's Avatar
    Join Date
    Jul 2015
    Location
    Gridania
    Posts
    1,566
    Character
    Song Sparrow
    World
    Balmung
    Main Class
    Goldsmith Lv 90
    Nobody ever remembers when they get 3 Excellent procs in a craft.
    (1)

    cerise leclaire
    (bad omnicrafter & terrible astrologian)

  9. #9
    Player
    Brightamethyst's Avatar
    Join Date
    Jul 2014
    Posts
    1,715
    Character
    Jenna Starsong
    World
    Goblin
    Main Class
    Scholar Lv 90
    Quote Originally Posted by MistakeNot View Post
    Unless you have several hundred tries (as an absolute minimum) to compare against, you have no idea if the RNG is good or not.
    You would need many thousands of samples to be able to tell for certain if it is working correctly.
    Only sort of. In the long term over hundreds or thousands of synths I'd bet the numbers would be pretty close to the listed percentages. I don't think anyone is denying or debating that. The issue is the short term and how the game gets to that point.

    I don't know how the game calculates things but the RNG is this game is way more streaky than in pretty much any other game I've ever played (in everything, not just crafting.) I expect random things to come in streaks occasionally, random is random after all, but I do not expect it to happen all the time. It's like the game tries to balance percentages by keeping track and swinging the pendulum back the other way if it gets too far off balance. That keeps the numbers straight over time but in insane in the short term. (Three 90% fails in a row, twice, in 20 minutes of crafting, should be a rarity, not a regular occurrence.)
    (0)

  10. #10
    Player
    TattersailNightwish's Avatar
    Join Date
    Sep 2018
    Posts
    142
    Character
    Tattersail Nightwish
    World
    Mateus
    Main Class
    Weaver Lv 90
    If I didn't know better (and I probably don't) I'd almost make a guess that when condition is Excellent, that some under-the-covers modifier gets baked into the equation.
    I would venture to say that I have a <=50% success rate on Synths when Excellent. I see "Excellent" and say "oh boy.." cross my fingers...

    Just my observation, YMMV. Just seems that way to me.
    (0)

Page 1 of 3 1 2 3 LastLast