Item Search
     
BG-Wiki Search
+ Reply to Thread
Page 3 of 6 FirstFirst 1 2 3 4 5 ... LastLast
Results 41 to 60 of 105

Thread: Kitron Macaron     submit to reddit submit to twitter

  1. #41
    E. Body
    Join Date
    Dec 2008
    Posts
    2,408
    BG Level
    7
    FFXI Server
    Quetzalcoatl

    https://spreadsheets.google.com/ccc?...FUnBNNGc&hl=en

    Just did another big batch, the two are slowly averaging out towards each other actually.

    However, if I take out day 2 as a bad sample (8% HQ on T4) then the macaron is still clearly up high. Time will tell.

  2. #42

    I found some chocolate rusks to mess around with and had 30 minutes before work. Made black ink (calamary and neb version, level 4 alchemy) and my alchemy is 55. No control group due to lack of time and fish, so this is just to throw info into the pot, can't really draw conclusions.

    24 synths
    2 breaks (lost 1 neb)
    12 NQ
    10 HQ (9 HQ1, 1 HQ2)

    Random details that likely don't matter: 21 synths done facing north in mog house, 3 synths done facing north east in windy waters while sacrificing a chicken to HQ gods. (Eatting a drumstick and last 3 synths were break with lost neb, 1 HQ1, and 1 NQ.) 60% moon.

    I have about 15 stacks of black ink stored up thanks to marlins being the fish ranking fish a week ago, so I'll mass produce some shihei and give some better tests when I can.

  3. #43
    Fishing Guru
    Join Date
    Jan 2007
    Posts
    4,723
    BG Level
    7

    Quote Originally Posted by urat View Post
    However, if I take out day 2 as a bad sample (8% HQ on T4) then the macaron is still clearly up high. Time will tell.
    isn't this the point of a large sample size? The good an bad days will even out thus cancelling the stretches people observe and then create theories based on. You can't go removing certain batches because they didn't meet a % you EXPECTED. That's pretty much the definition of being bias and skewing your results.

    Lastly what is t4? There's t-1(debatable), t0, t1, t2, t3.

  4. #44
    E. Body
    Join Date
    Dec 2008
    Posts
    2,408
    BG Level
    7
    FFXI Server
    Quetzalcoatl

    Quote Originally Posted by cdgreg View Post
    isn't this the point of a large sample size? The good an bad days will even out thus cancelling the stretches people observe and then create theories based on. You can't go removing certain batches because they didn't meet a % you EXPECTED. That's pretty much the definition of being bias and skewing your results.

    Lastly what is t4? There's t-1(debatable), t0, t1, t2, t3.
    T3* sorry

    Also in some scientific practices when taking multiple samples, if one sample is obscenely off from the rest, IE I'm getting 45%, 50%, 65%, 45%, 33%, 8%, a scientist assumes something went wrong with that last sample (Air escaped out of a tube, something contaminated a sample, etc) and simply takes it out.

    But thats when your individual samples are in extremely high sizes, like in the accuracies of .1 percents, not my 3% accuracy. So I don't think my sample size is large enough to reach the criteria to separate "bad" samples.

    Theoretically the reasoning is if I keep a bad sample and do 10,000 other samples that are all good, then the bad sample simply disappears under the sheer immensity of all the other samples that were actually the proper amount.

  5. #45
    New Merits
    Join Date
    Apr 2008
    Posts
    247
    BG Level
    4
    FFXI Server
    Bismarck

    How can you get a 'bad' sample in a video game though? Did you breathe on the text box and contaminate it with bactieria? Did something hit you in the face and you misread if you HQ'd or not? There is no random error with the game, if you put something in, you will always get the same thing out, its programmed that way. All the data has to be included.

  6. #46
    E. Body
    Join Date
    Dec 2008
    Posts
    2,408
    BG Level
    7
    FFXI Server
    Quetzalcoatl

    Quote Originally Posted by Jimothy View Post
    How can you get a 'bad' sample in a video game though? Did you breathe on the text box and contaminate it with bactieria? Did something hit you in the face and you misread if you HQ'd or not? There is no random error with the game, if you put something in, you will always get the same thing out, its programmed that way. All the data has to be included.
    Its mostly just a small sample size is the issue here, not a contamination.

  7. #47

    More quick testing with chocolate rusks because I may as well use the them. Not as many tests as I'd like, but I have to go to work rather early for the holiday weekend. Synth this time was shihei, level 93 wood, didn't bother with wearing GP items or support.

    Chocolate rusk - 36 synths - Wind day
    NQ: 18
    HQ1: 11
    HQ2: 4
    Break: 3 (2 bast, ink + 1 bast, 1 bast)

    No food - 26 synths - ice day/thunder day
    NQ: 16
    HQ1: 7
    HQ2: 2
    Break: 1 (ink)

    Rusks has promise, but not getting my hopes up. Still far too few tests to tell anything.

  8. #48
    E. Body
    Join Date
    Dec 2008
    Posts
    2,408
    BG Level
    7
    FFXI Server
    Quetzalcoatl

    Well you can't really compare doing batches of synths when one is on the considered "HQ" day and the other on the "success" day, kinda skews results XP

    I generally do 30~35 synths without food from 1:00 to 12:00, then 30~35 synths with food from 12:00 to 24:00 or as close as I can get, same direction, same craft, etc.

  9. #49
    Fishing Guru
    Join Date
    Jan 2007
    Posts
    4,723
    BG Level
    7

    Quote Originally Posted by urat View Post
    Well you can't really compare doing batches of synths when one is on the considered "HQ" day and the other on the "success" day, kinda skews results XP
    Here we go again...

  10. #50
    E. Body
    Join Date
    Dec 2008
    Posts
    2,408
    BG Level
    7
    FFXI Server
    Quetzalcoatl

    Quote Originally Posted by cdgreg View Post
    Here we go again...
    Whether or not it exists is null, my previous tests even showed there is likely little influence but as proper scientific study depicts, leave no variable unchecked.

    Especially when its not that hard to exclude it.

  11. #51

  12. #52
    Puppetmaster
    Join Date
    Jul 2007
    Posts
    52
    BG Level
    2

    It doesn't really matter when you guys are doing tests of 20-40 synths...

    Since I've never really done any testing in FFXI, what would be a sample that most people would respect?

  13. #53
    E. Body
    Join Date
    Dec 2008
    Posts
    2,408
    BG Level
    7
    FFXI Server
    Quetzalcoatl

    If you follow the procedure of my graph on the website, 30 or so synths without food and 30 synths with food, same direction same day same moon phase, would be very grateful for.

    Even 20 of each helps. So far I'm up almost up to a 100/100 sample size, which will put my accuracy to 1%, which is when the HQ rates start to become set in stone unless there are some really crazy days which might change it up or down by .5~1%

    But all these samples of with just food, or on different days, different moon phases, etc, are pretty much unusable for my sampling. =/

    It only takes 50 mins to do, and cuts out all but the one unknown variable, nonetheless I do appreciate your help, afterwards we can compile the rest and see if their hq rates went up by a set % IE 15%, 30->45, or if it just bumps your HQ rate up 1 tier or something like that, or grants you +skill, who knows.

  14. #54

    I'll play along as far as marking day/moonphase/direction, but I really don't care about what day it is, direction I'm facing, or the moonphase. I'm not really crafting to test this out, I'm just doing it whenever I realize I'm low on shihei (or ammo, but I haven't made any yet) and feel modivated enough to make more. The testing just takes an extra couple of minutes of my time.

    I'm doing chocolate rusks which seem less popular than coconut rusks anyway, based on sale history. So I'll probably have to settle for whatever the majority of my synths say, if anything. I'm operating on the hope that it's HQ% up, but I'm not really expecting anything amazing.

  15. #55
    Salvage Bans
    Join Date
    Mar 2006
    Posts
    749
    BG Level
    5
    FFXI Server
    Asura

    We are assuming Coconut Rusks are "stronger" than the others due to the recipe level, so it will be easier to see an impact if there is any.

  16. #56
    New Spam Forum
    Join Date
    Jan 2009
    Posts
    191
    BG Level
    3

    Did anything ever come of the desynth tests with the Kitron Macaron?

  17. #57
    Salvage Bans
    Join Date
    Mar 2006
    Posts
    749
    BG Level
    5
    FFXI Server
    Asura

    I am pretty convinced Kitron Macaron works.
    201 elemental ores to gems and only 1 break with Macaron and the NQ-only ring.

    Edit: Nvm, didn't notice the "desynth" part.

  18. #58
    Puppetmaster
    Join Date
    Nov 2008
    Posts
    62
    BG Level
    2
    FFXI Server
    Bahamut

    I used Kitron Macaron today while skilling up on a recipe that was 11 skills higher than my skill.

    I got quite a few 0.3 skill ups, I have never seen a 0.3 skill up before now.

    I'm not sure if this is because of the rather large level jump or the food.

    Is a 0.3 skill up possible and I'm just unlucky that I've never seen one before?

  19. #59
    Fake Numbers
    Join Date
    Nov 2008
    Posts
    81
    BG Level
    2
    FFXI Server
    Lakshmi

    You need to be pretty far from the skillcap to get .3's. You probably won't see any 5 levels or closer to the synth cap.

  20. #60
    Puppetmaster
    Join Date
    Nov 2008
    Posts
    62
    BG Level
    2
    FFXI Server
    Bahamut

    Quote Originally Posted by dem0nfr0g View Post
    You need to be pretty far from the skillcap to get .3's. You probably won't see any 5 levels or closer to the synth cap.
    Yea, as soon as I was within 9 levels of the cap the 0.3s disappeared.

    Good to know, thanks for the info.

+ Reply to Thread
Page 3 of 6 FirstFirst 1 2 3 4 5 ... LastLast