test content
What is the Arc Client?
Install Arc

upgrading artefacts, success 90%?

schietindebuxschietindebux Member Posts: 4,292 Arc User
edited March 2015 in General Discussion (PC)
i know its not a hot topic, but since I recognized it being a very very frequent error ingame i have to post it.
leveling artefacts from green to blue has a 90% success as its written, in my case that is abysmal wrong.
doing level up and failing two times (1% chance to happen) and if most level up fails first time very frequent i don´t believe these 90% at all because i believe in math and cryptics math seems to be wrong for sure
Post edited by schietindebux on

Comments

  • edited March 2015
    This content has been removed.
  • abaddon523abaddon523 Member Posts: 0 Arc User
    edited March 2015
    Ummm...his math isn't wrong. It's basic statistics. For something to have a 90% chance of success and then fail twice that has a (0.1*0.1) = 1% chance of happening.
    [SIGPIC][/SIGPIC]
  • schietindebuxschietindebux Member Posts: 4,292 Arc User
    edited March 2015
    Bernoulli can´t be that wrong
  • edited March 2015
    This content has been removed.
  • abaddon523abaddon523 Member Posts: 0 Arc User
    edited March 2015
    magenubbie wrote: »
    there's a big difference between statistics and chance.

    No. No there isn't. Probability quantifies chance.
    [SIGPIC][/SIGPIC]
  • schietindebuxschietindebux Member Posts: 4,292 Arc User
    edited March 2015
    ok then give me the formular for this task
    what is the chance of rolling two times a 1 out of two tries, rolling with a cube that has 10 sides?
    long ago i went to school btw
  • edited March 2015
    This content has been removed.
  • abaddon523abaddon523 Member Posts: 0 Arc User
    edited March 2015
    Anyway,

    If 10,000 people are upgrading artifacts at 90% probability of success the most likely distribution will be:

    9,000 people who succeed on the 1st try
    900 people who succeed on the 2nd try
    90 people who succeed on the 3rd try
    9 people who succeed on the 4th try
    and one person who gets really, really ticked and goes on to the message board to complain about how wrong the probabilities are being calculated by cryptic.

    It's quite possible you are are just one of the unlucky 100 who didn't get it on either of the first two tries.
    [SIGPIC][/SIGPIC]
  • schietindebuxschietindebux Member Posts: 4,292 Arc User
    edited March 2015
    no thats not the point, it happans nearly every time i level up, first fails almost allways...
    its not a real complaint , its more or less a bug imo, but i can be wrong, sure
  • valencayvalencay Member Posts: 431 Bounty Hunter
    edited March 2015
    yeah like when you need 30 pres wards for a 60% chance, bad luck? nah broken game
    Aset Xharran, Oppressor CW
  • abaddon523abaddon523 Member Posts: 0 Arc User
    edited March 2015
    no thats not the point, it happans nearly every time i level up, first fails almost allways...
    its not a real complaint , its more or less a bug imo, but i can be wrong, sure

    Given the number of bugs I've found in this game it wouldn't surprise me if they screwed something up here. But in order to draw that conclusion we'd need a much bigger sample size than your personal experiences in order for it to be statistically significant. Without a lot of people collaborating (or one person spending a ton of time on the preview server testing this) and posting their numbers it's just not possible to know whether the coding is flawed or you are just unlucky.
    [SIGPIC][/SIGPIC]
  • schietindebuxschietindebux Member Posts: 4,292 Arc User
    edited March 2015
    i don´t mind if i am not the one that is lucky, and i only wanted to post that there might be somthing wrong (as if that would be a big surpirse in this game)
    and sure i am the last one to test 100 hours on testserver for getting more numbers
    its a buggy game, tahst all, and probably this thread is old and fade since 100 ppl posted same issue before
  • namelesshero347namelesshero347 Member Posts: 2,109 Arc User
    edited March 2015
    People post only when they hit a nasty streak. No one post when they hit on the first attempt, even on a 10% chance.
  • valwrynvalwryn Member, NW M9 Playtest Posts: 1,620 Arc User
    edited March 2015
    Fey Blessings and Dragon Hoards hit at 1% of the time. If there wasn't a cooldown, they'd be triggering more often. This is not just a Cryptic thing, it's a D&D game thing. Neverwinter Nights was notorious for this. Everything was decided by a 20-sided die, yet 1's and 20's would count for 40% of the rolls (proven by many). Certain numbers are destined to be triggered more often than they should. :eek:
  • drezzatdrezzat Member Posts: 291 Arc User
    edited March 2015
    abaddon523 wrote: »
    Ummm...his math isn't wrong. It's basic statistics. For something to have a 90% chance of success and then fail twice that has a (0.1*0.1) = 1% chance of happening.

    Yeah I think he's right, well both of you are, you're justtalking about 2 different things , one doing the math and one working the probability math
  • juggalotus42000juggalotus42000 Member, NW M9 Playtest Posts: 47 Arc User
    edited March 2015
    yea its deff rigged to make u spend more $ or ad but to be fair you could try to upgrade said artifact a million times and fail every time because its not 100 percent
  • vaelynxvaelynx Member Posts: 182 Arc User
    edited March 2015
    drezzat wrote: »
    Yeah I think he's right, well both of you are, you're justtalking about 2 different things , one doing the math and one working the probability math

    yeah, no.
    Probabilities for independent events multiply, so the 0.01/1% is right on the money.
    However what is also right is that given the number of players, it's bound to happen to someone. Kinda like you sometimes get three 20s in a row in DnD which is even less likely but regardless I have seen it once.

    So yes, without gathering stats for a larger amount of players, it's hardly meaningful to talk about anything here.
  • duryntedurynte Member Posts: 132 Arc User
    edited March 2015
    Hi there!

    Talking of chance and random numbers, I would like to
    add a few thoughts on the subject. For a starter, random
    numbers produced by computing devices are, not without
    reason, sometimes called pseudo random numbers.
    http://en.wikipedia.org/wiki/Pseudorandomness
    If you happen to session on a unix box, you can take a
    look at those numbers via

    more /dev/urandom | od -tu1 -An # press ctrl-c to stop

    The device file /dev/urandom produces a sequence of bytes.
    There are advanced tests for judging on the quality of PRNGs
    (pseudo random number generators), especially on their
    distribution.

    Now, putting that in context, when I started to enjoy NWO
    just recently, I found myself puzzled with one particular
    event. Many of you may not remember your first character,
    just a tiny bit of gold available, so would be careful
    where to spend it. On potions? On skill node kits?
    I decided to gave the latter a try and looted skill nodes.
    And then I was surprized to find that, even though an about
    70% chance of success was stated, I usually failed, when I
    failed, about 3 times in a row.

    Going back to /dev/urandom. Take a sample of the sequence,
    devide the numbers by 2.55 (so they are ranged in 0-100),
    and take the first order difference sequence b_i = a_j - a_i.
    This is a distance messure, and in the samples taken distance
    was more often short than long. Translated back into NWO events
    the advice would be: If a chance based action fails, do not
    try again immediately (distance short = you will fail again).
    And that was just what I did on my first skill node fails.
    I would try loot (upgrade,..) but once, and never twice or
    more in a row.

    As to cut short rightfully raised objections to the above:
    As far as I can tell, every chance based action is PRNG'd
    at server side, which leaves the above vulnerable to some
    (not be answered) questions.

    So long, happy hunting!
    And may you all be PRNG-blessed by Thymora!
    (http://en.wikipedia.org/wiki/Thumos)
  • thesensaithesensai Member Posts: 637 Arc User
    edited March 2015
    drezzat wrote: »
    Yeah I think he's right, well both of you are, you're justtalking about 2 different things , one doing the math and one working the probability math

    Everyone is wrong, here's why:

    while, yes what you describe is accurate bases on true random probability, there is in fact NOTHING that is truly randomin computers or computer games. its just not possible to code a truly random number generator. It calculates the number based on a number of varying numbers, and changing variables. it might use a number in the internal game clock, coordinates you are on the map, maybe even what class you are on at the time, who knows?

    Point is, if you are having bad luck, do something to change your situation. log out for 15 min, zone somewhere else. Change your situation in as many ways as you can before you try again. I have actually heard this will even break unlucky streaks with lockboxes.

    Good luck!
  • edited March 2015
    This content has been removed.
  • cayappcayapp Member Posts: 826 Arc User
    edited March 2015
    To everyone who says broken...go play another game.
    Personally I have had 20% chance suceed when I forgot to put P ward in
    I have also had it fail 30 times in row with P wards.

    RNG monster ...if you know how to defeat please share.
  • thesensaithesensai Member Posts: 637 Arc User
    edited March 2015
    ihaveahies wrote: »
    The thing is that humans look for patterns that make not exist. That's why we have silly superstitions.

    Its not superstition, this is actually how they generate a 'random' number. There is nothing truly random in programming. But go ahead and troll if you must. lol
Sign In or Register to comment.