1. Hello!

    First of all, welcome to MapleLegends! You are currently viewing the forums as a guest, so you can only view the first post of every topic. We highly recommend registering so you can be part of our community.

    By registering to our forums you can introduce yourself and make your first friends, talk in the shoutbox, contribute, and much more!

    This process only takes a few minutes and you can always decide to lurk even after!

    - MapleLegends Administration-
  2. Experiencing disconnecting after inserting your login info? Make sure you are on the latest MapleLegends version. The current latest version is found by clicking here.
    Dismiss Notice

Not a Bug Taunt Debuff Not Increasing Drop Rate of Mobs

Discussion in 'Report Bugs' started by ChickenMan, Mar 28, 2019.

Thread Status:
Not open for further replies.
  1. OP
    OP
    ChickenMan
    Offline

    ChickenMan Mushmom

    55
    14
    55
    Dec 19, 2017
    Male
    6:41 AM
    MadMan
    Chief Bandit
    111
    OnDrugs
    LionheartLionheart , aaronisaaronis Kk, I'm going to show you guys a test I just did, so you know where I'm comming from. In this video I grind on the Dark Wyverns in the Canyon: West Road map while measuring my experience rate with the "epm #" command in the span of 3 minutes both while using and not using Taunt. Note that it takes me 2-3 hits(excluding the hit with the Taunt skill) to kill these mobs while I'm on the Coke Zero Pill while both using and not using Taunt. Sorry for the lack of audio.

    As you can see, the experience earned while using Taunt is significantly bigger than when you don't use it as is to be expected.
     
  2. KurayamiLove
    Offline

    KurayamiLove Skelegon

    951
    753
    413
    Mar 15, 2015
    Male
    Nautilus
    12:41 PM
    KurayamiLove/Nagrom/Amatista
    Buccaneer
    69
    Halcyon
    taunt is weird, also its hard to test the droprate as well because of teh RGN factor, another thing is that taunt is a bit slow (takes over a second to cast) so in the long run is decreasing the amont of mobs you kill per hour, for exp is worth but im not sure about droprate o.o. I think is all down for GMs to take a look at how it is codewise
     
  3. Krauser94
    Offline

    Krauser94 Designer Staff Member Graphics Designer

    387
    359
    274
    Nov 6, 2016
    Male
    12:41 PM
    Krauser94
    Shadower
    177
    Nebula
    Those 2 quotes sum up the main value of Taunt.

    Afaik, aaronisaaronis is aware of the exp since the start (and he is a fellow shadower too, so im pretty sure he knows how decent the exp is) so im sure he lived what you showed on the video a million times already SlimeXD
    What ppl is trying to say is that Taunt buffs Exp, CHANCE of drops and mob def by the same %.

    So basically, while you can test exp easily since the mob exp wont change, for drops is slightly more difficult to be fully sure it does since its based on RNG ("Random Number Generation"). Thats what they were trying to explain. You can kill 3 drakes and get 3 dragon skins in a row but you can kill 40 and get none for example. There is the random factor. As long as you can kill them near as fast using taunt than without it, its a win for the exp factor.

    And the "40% drop chance" as they said doesnt mean it raises the item falling to an OP number, more like it adds to the current multiplier, helping the drop "slightly"
     
  4. OP
    OP
    ChickenMan
    Offline

    ChickenMan Mushmom

    55
    14
    55
    Dec 19, 2017
    Male
    6:41 AM
    MadMan
    Chief Bandit
    111
    OnDrugs
    I'm sorry, but I don't get this. Why would it be worth it for the exp rate, but not the drop's?
     
  5. Krauser94
    Offline

    Krauser94 Designer Staff Member Graphics Designer

    387
    359
    274
    Nov 6, 2016
    Male
    12:41 PM
    Krauser94
    Shadower
    177
    Nebula
    Because lets say you kill a mob and it gives you 10k exp. If you kill the mob you ALWAYS get the exp, but you dont get always the drops, right? thats what i mean with the RNG issue. If you can kill a mob with one hit (one bstep) without using Taunt, but you need 2 hits after using taunt (lets say you do taunt-bstep-bstep), then its not that worth, because you could kill more mobs in the same time.

    What ppl means is, that as long the hits it takes you to kill the mob are the same either using taunt or not, the skill is totally worth, and really beneficial. But if you kill half of the mobs because it takes you more, even with that drop chance buff, the items dropping will be less, no matter what (less kills=less items). That combined with that RNG (again) makes the item dropping less noticeable.

    The other thing that is mentioned is the fact of killing low lvl mobs with it. Obviously as a shad you are going to kill, lets say a slime, in one hit even doing a normal attack, so if you are looking for his normal drops, you can mostly kill slimes with taunt and maybe see the effect more noticeable, since he is going to die in 1 attack always.

    (Im trying my best to explain it, sorry if im not clear enough SlimeCry)
     
  6. KurayamiLove
    Offline

    KurayamiLove Skelegon

    951
    753
    413
    Mar 15, 2015
    Male
    Nautilus
    12:41 PM
    KurayamiLove/Nagrom/Amatista
    Buccaneer
    69
    Halcyon
    Also I think the more mobs the skill can hit the less drop rate? o.o lots of variables to test this stuff x-x
     
  7. OhDoggo
    Offline

    OhDoggo Selkie Jr. Retired Staff

    204
    99
    179
    Jun 25, 2017
    Male
    12:41 AM
    Bowmaster
    This is indeed not a bug and the drop rate is working correctly for the skill. The reason why you think it is bugged has already been given the correct reasoning, but I will reiterate the main points.

    1) As others in the thread have mentioned or implied, the 40% drop rate increase works Multiplicatively. If a specific item from a mob has a chance to drop of say 0.01, and you taunt a mob, then the true drop value of the item becomes 0.01 x 1.4 = 0.014. Obviously some items may have much higher rates but these rates are not and will not be disclosed. If other servers do provide this, do not use that as a balancing pivot due to the fact how differently each server may edit their rates to be.

    2) RNG (Random Number Generator). It's a very simple concept that exists in practically all games that have a random drop/loot/gamble system implemented. Let's use a real life probability example: A 20 sided die has a 5% chance/ 1/20 chance to roll a certain number. Numerically, if you roll the die 100 times and count how many times it lands on '10', then you
    should roll 10 about 5 times. In reality, this is not the case due to this number just being a conclusion of theoretical probability. If you got lucky and rolled 10, 5 times in less than 50 rolls, who's to say that you're suddenly not allowed to roll 10 again despite having already rolled it 5 times? No one.

    3) Human based experiments/testing in games for topics like these, generally conclude inaccurate, unless you just happen to have that massive a sample size with consistent variables in each new test (The time spent, mobs killed, delay between killing x mob, etc...). There are exhaustively many variables that you just cannot keep consistent with each test you do, and because of the inconsistencies that happen with human based testing, it just cannot yield truly accurate results. If you really happen to believe your hypothesis of 'Taunt debuff not increasing drop rate' is true, then please commit to extensive hours (similar to that of time spent with machine learning) of testing, providing the variables you used as well as how you measured each variable and how you controlled it.

    Because there is ongoing discussion, I will not lock this thread (yet); however, if you want to disprove me and conclude that the drop rate taunt is indeed broken, then please do what i've asked you to do and extensively test the skill, measure your variables and how you controlled them.
     
    • Like Like x 2
  8. OP
    OP
    ChickenMan
    Offline

    ChickenMan Mushmom

    55
    14
    55
    Dec 19, 2017
    Male
    6:41 AM
    MadMan
    Chief Bandit
    111
    OnDrugs
    Before I do this, may I ask you that you can give me all the server's files to run the game(excluding the more sensitive information), so that I can have a better chance of testing the drop rate in a more controlled enviorment?
     
    • Like Like x 1
  9. Precel
    Offline

    Precel Zakum Retired Staff

    1,797
    247
    460
    Nov 13, 2018
    Male
    7:41 AM
    Precel
    Priest
    100
    Spirit
    That's a big nope. Basically most of server files are sensitive. You don't need the server files to run the experiment. You can control the environment with the tools you have at the moment, following other people's suggestions.
     
  10. nanop33
    Offline

    nanop33 Capt. Latanica

    364
    301
    279
    Jun 29, 2015
    Male
    Canada
    5:41 AM
    Retroviridae
    Shadower
    190
    Beaters
    That's not how statistics works. Just because the expected number of drops per number of kills is a certain number doesn't mean you're guaranteed that amount. The expectation only appears as your number of trials becomes infinitely large.

    Say you have 2 rigged coins, one with 1% chance of heads and one with 1.4% chance of heads. If you do an infinite number of trials, only then will the 1.4% coin will be absolutely guaranteed to land on heads more often than the 1% coin as the two distributions would stop overlapping.

    The larger the difference between two distributions, the more distinguishable they will be. But because the base drop rates are so low already, a 40% increase is negligible. Drop rates are not comparable with exp because you're comparing a 40% increase in the mean of a distribution with a 40% increase in a fixed number.
     
    Last edited: Mar 29, 2019
  11. OP
    OP
    ChickenMan
    Offline

    ChickenMan Mushmom

    55
    14
    55
    Dec 19, 2017
    Male
    6:41 AM
    MadMan
    Chief Bandit
    111
    OnDrugs
    Then can you spawn me in a secluded part of the server with 1 respawning mob with a 72% drop rate on the Etc?
     
  12. nanop33
    Offline

    nanop33 Capt. Latanica

    364
    301
    279
    Jun 29, 2015
    Male
    Canada
    5:41 AM
    Retroviridae
    Shadower
    190
    Beaters
    Assuming drop rates can be treated as a binomial distribution (item either drops or it doesn't), you can run the following code in R and adjust the values as you'd like:

    Code:
    # Adjustable values
    drop.rate <- 0.5
    kills <- 1000
    num.trials <- c(100, 200, 300, 5000)
    hist.bar.size <- 25
    
    for (trial.size in num.trials) {
      base.dist <- rbinom(n = kills, size = trial.size, p = drop.rate)
      base.hist <- hist(base.dist, plot = F, n = hist.bar.size)
      taunt.dist <- rbinom(n = kills, size = trial.size, p = drop.rate * 1.4)
      taunt.hist <- hist(taunt.dist, plot = F, n = 20)
     
      plot(
        base.hist,
        xlim = c(
                 min(base.hist$breaks, taunt.hist$breaks),
                 max(base.hist$breaks, taunt.hist$breaks)
               ),
        ylim = c(
                 min(base.hist$counts, taunt.hist$counts),
                 max(base.hist$counts, taunt.hist$counts)
               ),
        main = paste(
                 "Base vs. Taunted Drop Rate:",
                 trial.size,
                 "trials of",
                 kills,
                 "kills."
               ),
        xlab = "Num Drops",
        col = "grey"
      )
      plot(taunt.hist, add = T, col = rgb(.1, .1, .5, .5))
    }
    
    
    Keep in mind this is assuming the drop rate is realistic. Your suggestion at 0.72 would push the taunted rate to past 1.
     
    Last edited: Mar 29, 2019
    • Great Work Great Work x 1
    • Useful Useful x 1
  13. OP
    OP
    ChickenMan
    Offline

    ChickenMan Mushmom

    55
    14
    55
    Dec 19, 2017
    Male
    6:41 AM
    MadMan
    Chief Bandit
    111
    OnDrugs
    I'm sorry. What is "R"?
     
  14. nanop33
    Offline

    nanop33 Capt. Latanica

    364
    301
    279
    Jun 29, 2015
    Male
    Canada
    5:41 AM
    Retroviridae
    Shadower
    190
    Beaters
    It's a programming language. If you don't have access to it on your local machine, you can run it in an online console like here: https://rdrr.io/snippets/
     
  15. OP
    OP
    ChickenMan
    Offline

    ChickenMan Mushmom

    55
    14
    55
    Dec 19, 2017
    Male
    6:41 AM
    MadMan
    Chief Bandit
    111
    OnDrugs
    I would like to run this test in game to insure that the drop rate isn't affected by a glitch that may be in the server's code.
     
  16. Kimmy
    Offline

    Kimmy Administrator Staff Member Administrator Game Moderator

    10,963
    10,998
    551
    Sep 9, 2014
    The Netherlands
    12:41 PM
    Moderator Post
    We are not going to give you our server sided which gives you access to do anything you want with it, just to prove the skill isn't bugged. Furthermore, the drops are in the database, so I would have to give you the drop database as well. Not going to happen.

    Also, this thread got its answer, so I am going close it.
     
    • Like Like x 1
Thread Status:
Not open for further replies.

Share This Page