Limb Damage Equation concept

12346»

Comments

  • The "service department" thing was, in fact, humour.

    I will withhold my request for a refund as of now, of course.  If everything gets fixed in a timely manner, like I said, I'll be pretty happy.  I'm trying to be as unbiased as I can, here.  I'd give anyone else - person OR company - the same courtesy, after all.
    [2:41:24 AM] Kenway: I bet you smell like evergreen trees and you could wrestle boreal mammals but they'd rather just cuddle you
  • Dorns works fine for me, dunno why you lot struggle with it!




    Penwize has cowardly forfeited the challenge to mortal combat issued by Atalkez.
  • I'll agree with @Kuy: I bought a perfect limb counter. It currently isn't, but if I receive one, I would be happy with it all the same.
  • KuyKuy
    edited February 2015
    @Ernam:

    After some testing with Zulah and Leviticus, we've come to a strange conclusion.  At first, we thought maybe my band was bugged (on recommendation by @Eld).  However, the numbers say that cannot be the case (both dwarf miners with 12 strength, Zulah was doing ~5% more damage than me, consistently).

    However, changing Manda to assume I have a level 3 band has given me far, far more accurate numbers.  There are a few cases of strange breaks, but they're way, way less.

    Edit: It seems CompassSlash is still off.  If I use legslash left to break, then bring the right leg within one off-slash via CompassSlash, the next legslash left does NOT break the right leg.
    [2:41:24 AM] Kenway: I bet you smell like evergreen trees and you could wrestle boreal mammals but they'd rather just cuddle you
  • BluefBluef Delos
    edited February 2015
    Ernam said:

    It is certainly worth noting that both of these individuals were banned from speaking to/about me for doing the exact same thing in countless other threads, and apparently just decided to (temporarily?) ignore that fact.

    If you're referring to me, which I can only assume you are since you're responding to Grandue, you're lying. I have never been forum banned. I have never been banned from speaking to or about you on any other thread.

    I won't de-rail the thread further though since you seem to have gotten a halfway reasonable post written here, finally.
  • I just don't understand how Sarapis's own limb damage equation doesn't hold up on his own game. Kuy, you should probably file a detailed bug report.

  • NimNim
    edited February 2015
    Report part 1.5: I wrote a script that uses Manda to print out csv-formatted break counts as per my previous data set. This'll make comparing outputs easier. If Ernam'd like it, I can send it to him, naturally, for his scrutiny or other purposes, once I clean it up. :3
  • To quote myself from another thread where you made a claim, supporting it as 100% accurate, and it turned out that it wasn't actually 100% accurate:

    Jarrod said:
    Do yourself and everyone a favor and don't state things you can't support as facts.

    If you had advertised your equation as: "Accurate through 3,000+ tests across extreme variations in health and limb attacks." and "If anyone does find errors at any data points, I will update accordingly, though it is unlikely they exist." There would probably still be people trying to troll you in the thread, that's basically an inevitability at this point, but you would also have people supporting you too.

    I'm not saying don't try to sell your system, I'm not saying you're a bad coder (you're obviously not), I'm saying your approach to the situation is wrong. Realize that everyone, yourself included, can make mistakes and then advertise like that, and you'll have a lot more success.
    image
    Cascades of quicksilver light streak across the firmament as the celestial voice of Ourania intones, "Oh Jarrod..."

  • NimNim
    edited February 2015
    Part 2:

    My previously mentioned assumption that Arash > Doya > Sanya > Unstanced > Mir > Thyr has been personally re-confirmed, at least for the tested value of 8656 HP. I'm assuming that this is generally true, as my old data suggested it's true as well, but never trust a single data point, kids. Limbprobe was used extensively in this testing. @Silvarien is pretty cool sometimes.

    The problem is that Manda assumes (Arash = Doya) > Unstanced > Sanya > Mir > Thyr. This is the first thing I'd address.

    Note that @Ernam has not sent me the actual formula he derived for Manda - I only have its source code and output to go by, so anything found in these reports could just be code bugs, oversights, etc. I wouldn't know, but naturally, if I receive an updated version, I'll amend any new findings that result from it. :3

    The other problem is that the test results for 8656 HP matched Manda's output six times out of eighteen across all six stances and all three attack options available in Manda, with a level 1 band.

    Test results: http://pastebin.com/s46RcSFf
  • just an FYI on dorns, if you have a level 2 band, up the band to level 3 in the limb counter code and it breaks correctly.
  • And Manda has the exact same issue and solution?
    image
  • Rangor said:
    And Manda has the exact same issue and solution?
    No, it does not.
  • Rangor said:
    And Manda has the exact same issue and solution?

    Unless Dorn's limb tracker changed from what I remember and what google could find, the two use very different formula. Dorn's tracker also uses "hand-picked" damage values that he got through testing, while Manda attempts at a standard formula.

    It's still possible that the issues are similar, but it'd be a coincidence or at best resulting from similar tests / testing methods.


  • In the event that Manda's current BM equation is slightly off, it certainly would not show often.  If you're seeing frequent calculation errors, there is a bigger problem.  The actual limb damage equation may have changed, there could be a setting issue, among a few other possible causes.

    Going to go back into BM as soon as SnB is wrapped up to investigate this issue.
  • The formula could have changed, yes. That, or you didn't have the exact formula after all.

    Probably one of those two.

  • After kuys testing it looks more like an issue with band calculation. can't do it like you do knuckles for monks. 

  • edited February 2015
    Cooper said:
    After kuys testing it looks more like an issue with band calculation. can't do it like you do knuckles for monks. 
    I did actually verify with @Makarios That it is a flat 5/10/15% multiplier, which is what Help Artefacts states.  It's also what Manda uses, and I verified (at the time) that all three tested out correctly (particularly since I own one).

    Again, will be looking into this ASAP.
  • Cooper said:
    After kuys testing it looks more like an issue with band calculation. can't do it like you do knuckles for monks. 
    Kuy also only tested in one stance, though, and the one with the largest adjustment to base damage, to boot. Stances and bands seem to affect the damage in the same way, so changing the band setting could also be compensating for an error in the stance calculation. Or there could be something weird going on with the combination of the two. Plenty of possibilities.

    Ernam said:
    Cooper said:
    After kuys testing it looks more like an issue with band calculation. can't do it like you do knuckles for monks. 
    I did actually verify with @Makarios That it is a flat 5/10/15% multiplier, which is what Help Artefacts states.  It's also what Manda uses, and I verified (at the time) that all three tested out correctly (particularly since I own one).

    Again, will be looking into this ASAP.
    I don't suppose he specified whether it adds int(damage*.05)*band_level or int(damage*.05*band_level)? Not that I think that difference is the source of any of what's being discussed, I've just been trying to figure out how the rounding is done for a while.
  • In case kuy didn't tell you, when he manually switched to level 3 band it was almost perfect. 

  • Cooper said:
    In case kuy didn't tell you, when he manually switched to level 3 band it was almost perfect. 
    I know. My point is that is that switching bands is equivalent to switching stances (for example, damage with a level 2 band in Thyr is the same as with no band in Mir), so it wouldn't be surprising to have an error in the stance calculation that was perfectly corrected by changing the band setting.
  • edited February 2015
    At the time I did the evaluations on leg/arm/centreslash, what Eld said appeared to be the case (both band and stance being flat multipliers), so what he's suggesting may be due to a change in stance multiplier or a bad multiplier for Thyr in Manda.  I tested everything unstanced, and later calculated out the stance mods based on a separate set of runs (to determine the mid amount), so it's very possible that the equation is still fine.

    Will confirm/fix as needed when I can get on a BM and troubleshoot.
Sign In or Register to comment.