Page 39 of 89 FirstFirst ... 29 37 38 39 40 41 49 ... LastLast
Results 381 to 390 of 885
  1. #381
    Player Insaniac's Avatar
    Join Date
    Mar 2011
    Posts
    1,003
    Character
    Insaniak
    World
    Lakshmi
    Main Class
    THF Lv 99
    My goal would be to prove within a reasonable margin of error that the TH3 trait is or isn't a reroll. Since it would be a 10% difference in drop rates what kind of sample size would I need for a 4% margin of error? Please forgive my math deficiency.

    edit: Considering the alternative would be that TH3 is just TH+1 we better make that a 3% margin of error.
    (0)
    Last edited by Insaniac; 08-06-2011 at 10:51 PM.

  2. #382
    Player
    Join Date
    Mar 2011
    Posts
    237
    I also don't know much on statistcs. Maybe i'm just saying shit here (so correct me, please~), but i think that if you parsed a 70% drop rate, and this sample had a 10% margin of error, we could very well conclude that TH3 is also a reroll. This is because the 11k parse is very solid. We would need lower margins of error if we didn't know TH2 so well.
    (0)

  3. #383
    Player Insaniac's Avatar
    Join Date
    Mar 2011
    Posts
    1,003
    Character
    Insaniak
    World
    Lakshmi
    Main Class
    THF Lv 99
    I'm just looking for the absolute minimum amount of effort to come up with a solid result. =P
    (0)

  4. #384
    Player Nebo's Avatar
    Join Date
    Apr 2011
    Location
    Thief
    Posts
    475
    Character
    Nebo
    World
    Lakshmi
    Main Class
    THF Lv 99
    Quote Originally Posted by Insaniac View Post
    I'm just looking for the absolute minimum amount of effort to come up with a solid result. =P
    That would be for SE to just *gasp* tell us what it does.

    I don't know why they think it is fun that we still don't know after all these years.

    This is not fun.
    (0)

  5. #385
    Player Byrth's Avatar
    Join Date
    Mar 2011
    Posts
    2,172
    Character
    Byrth
    World
    Lakshmi
    Main Class
    DNC Lv 99
    Approximately 90~100 kills should let you see the difference.
    (0)

  6. #386
    Player Insaniac's Avatar
    Join Date
    Mar 2011
    Posts
    1,003
    Character
    Insaniak
    World
    Lakshmi
    Main Class
    THF Lv 99
    Quote Originally Posted by Byrth View Post
    Approximately 90~100 kills should let you see the difference.
    Pfffft. Consider it done. I'll blast these down in the next couple days.
    (0)

  7. #387
    Player Byrth's Avatar
    Join Date
    Mar 2011
    Posts
    2,172
    Character
    Byrth
    World
    Lakshmi
    Main Class
    DNC Lv 99
    Heh, okay! Remember to note which ones ding up to TH level 4!

    Also, that was an approximation, the formula you're using is:

    95% Confidence interval width = 1.96*SQRT(p*(1-p)/N)
    Where p = Number of Feathers or Beaks / Number of Kills
    and N = Number of Kills

    So if you get 70/100 feathers, that's 1.96*SQRT(.7*.3/100) = +/-.09 (9%) for your 95% confidence interval width
    (0)
    Last edited by Byrth; 08-07-2011 at 02:18 AM.

  8. #388
    Player
    Join Date
    Mar 2011
    Posts
    237
    It's accepted that SAWS also triggers proc, right? Like, procs but doesn't show on log. Better not use stacked ws just to be safe as well.

    I would help you to speed things up but i've been very busy these days :\
    (0)
    Last edited by Laphine; 08-07-2011 at 02:46 AM.

  9. #389
    Player noodles355's Avatar
    Join Date
    Mar 2011
    Posts
    883
    non-stacked WS can proc too. You'd be best not WSing at all.
    (0)

  10. #390
    Player Insaniac's Avatar
    Join Date
    Mar 2011
    Posts
    1,003
    Character
    Insaniak
    World
    Lakshmi
    Main Class
    THF Lv 99
    Quote Originally Posted by noodles355 View Post
    non-stacked WS can proc too. You'd be best not WSing at all.
    Yeah that's not happening. I'll just kill a few hundred to offset the possible hidden TH procs.
    (0)

Page 39 of 89 FirstFirst ... 29 37 38 39 40 41 49 ... LastLast