Item Search
     
BG-Wiki Search
+ Reply to Thread
Page 1 of 11 1 2 3 ... LastLast
Results 1 to 20 of 212

Thread: Magnetic Earring Testing     submit to reddit submit to twitter submit to tumblr

  1. #1

    Magnetic Earring Testing

    http://tyladras.livejournal.com/117255.html

    Did some testing after I got my earring, posted the results on my LJ above - enjoy, hopefully it helps a few people out with their decisions ;3

  2. #2
    Physicist
    Join Date
    Feb 2005
    Posts
    4,493
    BG Level
    7
    FFXIV Character
    Raineer Severus
    FFXIV Server
    Hyperion
    FFXI Server
    Siren
    WoW Realm
    Area 52

    Decent test but not nearly enough datapoints :/

    I can promise you that it doesn't take the proc rate from 20% to 33% (since the typical activation rate over time is 25%)

  3. #3

    It's good to know it works without BLM sub, but your number of tests is too small to really say anything else.

  4. #4
    Physicist
    Join Date
    Feb 2005
    Posts
    4,493
    BG Level
    7
    FFXIV Character
    Raineer Severus
    FFXIV Server
    Hyperion
    FFXI Server
    Siren
    WoW Realm
    Area 52

    Quote Originally Posted by FFXIFlux
    It's good to know it works without BLM sub, but your number of tests is too small to really say anything else.
    agreed, definately is good to know that.

  5. #5
    Sea Torques
    Join Date
    Sep 2005
    Posts
    657
    BG Level
    5

    more data !

    nice start though.

  6. #6

    The main purpose was to change guesses to semi-conclusive figures. Ideally I would have performed the test on some additional spells, with 100 casts per. I don't know if I have the time to commit for it ^^

    Like I say, the general idea was to show the changes in approximate effectiveness - there is enough data to show that in what I've done already ;3

  7. #7

    Quote Originally Posted by Enchy
    Like I say, the general idea was to show the changes in approximate effectiveness - there is enough data to show that in what I've done already ;3
    No, there is really not enough data there to tell you anything, not even approximatly. You need a much larger sample.

  8. #8

    Quote Originally Posted by FFXIFlux
    Quote Originally Posted by Enchy
    Like I say, the general idea was to show the changes in approximate effectiveness - there is enough data to show that in what I've done already ;3
    No, there is really not enough data there to tell you anything, not even approximatly. You need a much larger sample.
    It tells you that the earring does not require a BLM subjob, and that based on the sample size, it does increase the proc rate, even if not by much.

    This was carried out as a personal study, more for my own information than anyone elses, I just decided to share because it indicates certain points.

    Obviously, the sample is not representative enough to make any concrete conclusions, but that wasn't the original intention of the short study. Basically, it requires further investigation, which is something I may do at a later date.

  9. #9
    Melee Summoner
    Join Date
    Oct 2005
    Posts
    39
    BG Level
    1

    Quote Originally Posted by FFXIFlux
    No, there is really not enough data there to tell you anything, not even approximatly. You need a much larger sample.
    Until you start quoting stats, or confidence levels, you can't talk intelligently about sample size.

  10. #10
    Yoshi P
    Join Date
    Jul 2004
    Posts
    5,352
    BG Level
    8
    FFXI Server
    Bahamut

    Enchy, don't pay attention to what Flux says. I was banned from these forums for about 5-6 months, when i came back recently, I was introduced to his words of "wisdom". From my general impression, he usually follows the crowd in his opinion and exmphasizes negative parts in order to make himself look cool. If you note, most of his opinions are looked at without any shread of respect and treated as a joke. :wink:

  11. #11
    Physicist
    Join Date
    Feb 2005
    Posts
    4,493
    BG Level
    7
    FFXIV Character
    Raineer Severus
    FFXIV Server
    Hyperion
    FFXI Server
    Siren
    WoW Realm
    Area 52

    Quote Originally Posted by qatsi2
    Quote Originally Posted by FFXIFlux
    No, there is really not enough data there to tell you anything, not even approximatly. You need a much larger sample.
    Until you start quoting stats, or confidence levels, you can't talk intelligently about sample size.
    Tests showing 1500+ casts put the activation rate right at around 25%. He got 20%, so his test is +/- 5% margin of error?

    I don't want to sound like i'm indicting the OP because I'm not...he's provided more data than anyone else I've seen with the earring. However you can't conclude anything from the test besides the fact that it does activate without /BLM (which was something that needed proving.)

  12. #12

    Quote Originally Posted by Enchy
    It tells you that the earring does not require a BLM subjob, and that based on the sample size, it does increase the proc rate, even if not by much.
    Yes, it tells you it works without BLM sub. NO, it DOES NOT tell you it increases proc rate. Your sample size is too small. While it suggests it, you are still within error rate at 60 samples and are not be able to make the claim that it increases proc rate.

    Quote Originally Posted by Enchy
    This was carried out as a personal study, more for my own information than anyone elses, I just decided to share because it indicates certain points.
    Thanks for figureing out it works without BLM sub, but you will want to do more testing to see how worthwhile it is with it. The fact that with the earring but not sub you get 4 procs and save almost as much MP as off 14 procs with blm sub but no earring should be some sort of indicator as to how accurate such a small sample is...

  13. #13

    Quote Originally Posted by qatsi2
    Quote Originally Posted by FFXIFlux
    No, there is really not enough data there to tell you anything, not even approximatly. You need a much larger sample.
    Until you start quoting stats, or confidence levels, you can't talk intelligently about sample size.
    http://www.causascientia.org/math_stat/ ... ionCI.html?

    Enjoy.

  14. #14

    Quote Originally Posted by Hirronimus
    Enchy, don't pay attention to what Flux says. I was banned from these forums for about 5-6 months
    That pretty much says what needed saying right there Hirroniums.

  15. #15

    Quote Originally Posted by FFXIFlux
    Quote Originally Posted by Enchy
    It tells you that the earring does not require a BLM subjob, and that based on the sample size, it does increase the proc rate, even if not by much.
    Yes, it tells you it works without BLM sub. NO, it DOES NOT tell you it increases proc rate. Your sample size is too small. While it suggests it, you are still within error rate at 60 samples and are not be able to make the claim that it increases proc rate.

    Quote Originally Posted by Enchy
    This was carried out as a personal study, more for my own information than anyone elses, I just decided to share because it indicates certain points.
    Thanks for figureing out it works without BLM sub, but you will want to do more testing to see how worthwhile it is with it. The fact that with the earring but not sub you get 4 procs and save almost as much MP as off 14 procs with blm sub but no earring should be some sort of indicator as to how accurate such a small sample is...
    I'm not denying the fact that the sample size needs to be larger, I've already told you what I would have set it at ideally had I the time to test it This is something that I'll be working on over time, because I'm not sitting down to the game for a whole day calculating MP use Obviously, I'll let you guys know when I update, but for the love of cookies understand that it is permissible to make assumptions on small data samples if you have the intention of further investigating to prove a hypothesis either true or false ;x

    EDIT: Anyway, if anyone's interested, the next steps in my 'study' involve using a single spell (Cure 3 - higher MP cost allows for a more accurate percentage to be formed rather than playing with x/8 we're looking at x/46 instead, as well as not requiring me to refresh my MP as frequently) with the same job/earring combinations as before, but with a larger sample size to reduce the margin of error that exists. This will further my hypothesis that the conserve MP level increases the proc change of the ability as well as it's potency, but I'll be looking specifically at proccing first.

    For now, be happy knowing that it doesn't require /BLM ;3

    Oh, and thanks to the people who have given me valuable input, and driving me to the point of wanting to test more - if only to stop a certain someone nagging about error margins and confidence intervals! <3

  16. #16
    The Optimistic Asshole
    Sweaty Dick Punching Enthusiast

    Join Date
    Aug 2005
    Posts
    23,172
    BG Level
    10
    FFXIV Character
    Tyche Six
    FFXIV Server
    Tonberry

    I ran tests of this a few days ago. Did it as WHM/WAR. Casted nothing but Cure 2 the entire time, proc'd 1:50 casts (on the 49th cast). Took Cure 2 from 24 MP to 21 MP. The earring is complete shit. We've been arguing over this earring for the past several days here:

    http://ffxi.killvoid.com/forums.php?m=posts&q=107844

    I stole the figures of conserve MP equations that aurik posted here somewhere in that thread. Hope he didnt mind.

  17. #17
    A Magic Ham Sandwich
    Join Date
    Nov 2005
    Posts
    5,388
    BG Level
    8

    For a non BLM or blm sub it might suck... but when I'm duoing with BLMs that use AF1 feet I notice that they often have much more MP when I'm already out.

    Sort of like Rostrum pumps I would guess. Still work without rdm sub, but only takes off 1 second of recast time so barely noticeable at all.

  18. #18

    Raineer made a good post today about BLM with Weskit and Water Ring (+2 and +15 conserve MP) http://www.theorderls.com/phpBB2/viewto ... &start=330

    With 300 casts the Weskit saved LESS MP than without... because even 300 isn't a very big sample when testing such a small amount. With +17 conserve MP a BLM went from 6.6% MP saved to 11.4%. How will +5 Conserve MP from the earring do? It's tough to say, but it's looking a lot like it will save you something like 1% MP.

  19. #19

    Quote Originally Posted by FFXIFlux
    Quote Originally Posted by Hirronimus
    Enchy, don't pay attention to what Flux says. I was banned from these forums for about 5-6 months
    That pretty much says what needed saying right there Hirroniums.
    Things shoulda stayed that way too.

    GTFO

  20. #20
    Yoshi P
    Join Date
    Jul 2004
    Posts
    5,352
    BG Level
    8
    FFXI Server
    Bahamut

    Geno still trying to be the Mod of BG forums? /angst /angst?

+ Reply to Thread
Page 1 of 11 1 2 3 ... LastLast

Quick Reply Quick Reply

  • Decrease Size
    Increase Size
  • Remove Text Formatting
  • Insert Link Insert Image Insert Video
  • Wrap [QUOTE] tags around selected text
  • Insert NSFW Tag
  • Insert Spoiler Tag

Similar Threads

  1. Paladins with Magnetic earring? (vs Ethereal Earring)
    By grimrising in forum FFXI: Everything
    Replies: 16
    Last Post: 2008-07-29, 09:04
  2. Need advice Ethereal Earring or Magnetic Earring ?
    By yoshizz in forum FFXI: Everything
    Replies: 152
    Last Post: 2006-06-07, 11:45
  3. Replies: 1
    Last Post: 2004-11-15, 19:53