Item Search
     
BG-Wiki Search
+ Reply to Thread
Results 1 to 5 of 5
  1. #1
    Sea Torques
    Join Date
    Feb 2005
    Posts
    699
    BG Level
    5
    FFXI Server
    Asura

    StoreTP Test - Regarding "Pre-Truncation" and TP/hit

    As a preface, I should say that this is my first time doing an in-game test of any real substance. I don't post here at BG often and this is my first time starting a new thread here. If the thread isn't BG-worthy simply cuz it's on such a narrow topic, or because the test is so simple, or because the test is shown to be flawed or wrong, or because the test has been done/proven before but the wiki(lol) hasn't yet been updated, my apologies. 8)

    What the test is about, and why I did it:The goal of this test is to get a more precise concept of how StoreTP affects TP/hit. What it's testing is whether or not the TP/hit value is truncated before the StoreTP multiplier, rather than only after StoreTP has been multiplied. Wiki(lol) says that the StoreTP is truncated only after the StoreTP multiplier...
    When calculating TP after accounting for Store TP, FFXI drops all decimal places after the first.
    ...but does not mention whether or not there is truncation before the multiplier, and it almost sounds as if its implying that there is no truncation prior to multiplying StoreTP. In any case, the reason for doing a test like this, as many would predict, is to provide more precise knowledge as to how effective certain StoreTP setups would be.

    How to do the test:
    1.Pick a weapon combination.
    2.Find out the combinations TP/hit value.
    3.Pick a certain amount of StoreTP.
    4.Figure out mathematically what the end TP/hit would be after factoring StoreTP (a)if TP/hit is truncated prior to the StoreTP multiplier, and (b)if TP/hit is only truncated after the StoreTP multiplier.
    5.Make sure that the difference between the hypothetical TP/hit values of (a) & (b) are such that they do not give the same end-result TP/hit after final truncation (otherwise the test would be pointless).
    6.Go out and do the test, tallying up the total # of times for each different TP-return. Then calculate the average to get your TP/hit result.

    The test itself: http://www.youtube.com/watch?v=uECWxzg7ekE
    (The method used in the test could have been better. I had thought before doing it that TP-return on melee hits worked on probability, IE you have a 10% 'chance' to hit for 4TP with 4.9TP/hit. I then found out that the TP-return worked on queue/cycle--not probability. Thus a ten-hit (or multiple there of) method would have been less sloppy. Also, only the first of the three test attempts would have been necessary to support the hypothesis, as TP/hit would have been 5.0, had the hypothesis been incorrect.)

    Initial thread(s):
    http://ffxi.allakhazam.com/db/jobs.h...91417121419340
    (I know, it's Allah'. Sue me.)

    Summary of results:
    Weapon Combo:Blau/Sirocco+Suppa, delay/hand 131.2 ---- from ((178+150)*0.8 )/2
    Tp/hit: 4.593 (trailing) ---- from 5.0+((131.2-180)*1.5)/180
    Truncated:4.50
    StoreTP+11 with 4.593TP/hit:5.098(or 5.0) ---- from 4.593*1.11
    StoreTP+11 with 4.5TP/hit:4.995(or 4.9) ---- from 4.5*1.11

    Test#1:StoreTP+11 - Carbonara + Rajas
    Hypothetical TP/hit if truncation occurs after the StoreTP multiplier: 5.0
    Hypothetical TP/hit if truncation occurs before the StoreTP multiplier: 4.9
    Actual TP/hit experienced upon testing: 4.868 & 4.889
    (I did this test twice). The first test was calculated from a total of 38 hits, and the second from a total of 45 hits.

    Test#2:StoreTP+12 - Carbonara + Rajas + Brutal
    Hypothetical TP/hit if truncation occurs after the StoreTP multiplier: 5.1
    Hypothetical TP/hit if truncation occurs before the StoreTP multiplier: 5.0
    Actual TP/hit experienced upon testing: 5.0
    (I never hit 4tp or 6tp once). Calculated from a total of 38 hits.

    Conclusion:TP/hit is truncated before the StoreTP multiplier. Then StoreTP is multiplied, and then it is truncated again.


    Now just get that BG wiki up so I can stop using this the gei one. Once again, sorry if this thread is thoroughly uninteresting, or if it confirms results that had already been confirmed by other players. The (soon to be the old) central source for detailed FFXI info, ffxiclopedia wiki, didn't include anything about pre-truncation in the article, and seemed as if to state the contrary. I'm sure there are others who already believe/know about the pre-truncation, but as of right now it lacked confirmation in wiki, and I could not find other tests like this one that had been done before. 8) Finally, if my results are in fact wrong in some way, let me know. I am thoroughly new at this sort of thing.

  2. #2
    Sea Torques
    Join Date
    Dec 2006
    Posts
    590
    BG Level
    5
    FFXIV Character
    Dense Fever
    FFXIV Server
    Ragnarok
    FFXI Server
    Bahamut

    Re: StoreTP Test - Regarding "Pre-Truncation" and TP/hit

    Just as a small point for testing, to easily figure the TP per hit of any weapon, simply use it through 10 hits and then divide the TP gained by 10. I find it much easier than predicting the mathematical decimal figure at certain times and with certain weapons.

  3. #3
    Sea Torques
    Join Date
    Feb 2005
    Posts
    699
    BG Level
    5
    FFXI Server
    Asura

    Re: StoreTP Test - Regarding "Pre-Truncation" and TP/hit

    (The method used in the test could have been better. I had thought before doing it that TP-return on melee hits worked on probability, IE you have a 10% 'chance' to hit for 4TP with 4.9TP/hit. I then found out that the TP-return worked on queue/cycle--not probability. Thus a ten-hit (or multiple there of) method would have been less sloppy...)
    Yay, one reply so far. Guess my test was every bit as unexciting as I expected, hehe. Yeah, on this, I had thought upon going into the test that TP hit return worked on probability. Afterwards it was brought to my attention in the other thread that it worked on a queue/cycle. SE never ceases to surprise me in the way that they chose math and stats to work in this game.

    If anything, I want to know from any of you guys if you had ever heard of StoreTP working like this before, and if you think you can use the results of make better gear choices for your job.

    For Thf, +12StoreTP to bring back the 5tp floor for that weapon combo isn't too shabby. Rajas+Brtual+Carbonara ftw. Also means Skadi Chausses + Rajas works. Unfortunately though it looks like there's no use in ever dropping Rajas for Toreador or Sniper+1 unless one is /war.

  4. #4

    Re: StoreTP Test - Regarding "Pre-Truncation" and TP/hit

    Yes, every step truncates decimals, it's a limitation of the representation of numbers they are using (particularly, they use integers and calculate TP in tenths (ie, 1 TP = 10 units), so every operation tosses away the decimal expansion).

    A better test in the future would be to use a 2H weapon or a single 1H weapon, so that you don't have to correct for dual-weild (which is a small source of error / risk).

  5. #5
    Banned.

    Join Date
    Oct 2006
    Posts
    10,159
    BG Level
    9

    Re: StoreTP Test - Regarding "Pre-Truncation" and TP/hit

    Quote Originally Posted by Shamaya
    (The method used in the test could have been better. I had thought before doing it that TP-return on melee hits worked on probability, IE you have a 10% 'chance' to hit for 4TP with 4.9TP/hit. I then found out that the TP-return worked on queue/cycle--not probability. Thus a ten-hit (or multiple there of) method would have been less sloppy...)
    Yay, one reply so far. Guess my test was every bit as unexciting as I expected, hehe.
    Actually a low reply count is good, it means everyone: a) agrees, b) is reading, c) has nothing to say.

    A bad post would be 2 pages long of flames. 8)

Similar Threads

  1. double attack and multi hit weapons
    By Raizins in forum FFXI: Everything
    Replies: 19
    Last Post: 2009-06-08, 14:33
  2. Promy bosses and TP moves
    By Vidar in forum FFXI: Everything
    Replies: 4
    Last Post: 2008-06-24, 07:35
  3. Gorgets and multi-hit accuracy
    By Valyana in forum FFXI: Everything
    Replies: 23
    Last Post: 2008-03-11, 22:55
  4. Zanshin and multi-hit weapons
    By Daggerot in forum FFXI: Everything
    Replies: 7
    Last Post: 2007-10-02, 13:24
  5. Bio and TP gain
    By Alastaire in forum FFXI: Everything
    Replies: 6
    Last Post: 2007-03-28, 13:36
  6. Tests regarding delay cap
    By Nny in forum FFXI: Everything
    Replies: 32
    Last Post: 2006-08-22, 02:19
  7. Tests Regarding Sushi
    By nivlakian in forum FFXI: Everything
    Replies: 69
    Last Post: 2006-07-21, 22:17