Demoralizing RNG Algorithm for 5* pulls

peanut_gallery
peanut_gallery Posts: 119 Tile Toppler
edited August 2017 in MPQ General Discussion
To quote Rod Stewart circa the 80's, "some guys have all the luck." And this season, apparently D3 has decided that I am not one of them.

Since 5* Spidey entered LL tokens, I've pulled 104 LL's, and only gotten 11 5* covers, giving me under an 11% pull rate on what is arguably a fairly statistically significant sample size.

I started out with a fairly amazing 9/30 pull rate, which was  unprecedented, and I knew that the odds would say that I would eventually balance out to around 15%.  Since then I have gone 2/74, which is just soul crushing when when it comes to incentive for progressing in the game (not mention that I went 1/28 for the 10% Spidey vault, which was also pretty demoralizing).

If D3 is truly quoting 15% odds for 5* covers, they need also specify within what number of pulls that rate applies for "the long run" and find a way to make good on that.  I'd gladly take a more balanced distribution (i.e. forgo my amazing 9/30 streak) so I wouldn't need to endure a 2/74 and counting cold streak.

Has anyone else experienced this craziness?  How has it impacted your will to stay with the game?
«1

Comments

  • jredd
    jredd Posts: 1,387 Chairperson of the Boards
    all i know is, as soon as i champed my o m l (2 weeks ago) all i see in pvp is champed 5*after champed 5* so some doods gotta be pulling them enough to champ them all.
  • CHRISJN
    CHRISJN Posts: 116 Tile Toppler
    Sorry about your "dreaded" pull, peanut.
    If like that, hoarding seems no use... it still luck based. Last 3 days i become itchy and used my 400cp... result was a bit positive though.
    1 black surfer - to fully cover him- championed
    1 thanos black
    1 Gg purple
    1 strange blue
    1 natasha red
    1 phx green
    1 black banner
    Thats 7/20 or 35%...
    Rest are all 4* latest pulls. Fully covered vulture, he is my next 4 to champ.
    Again its just pure luck...


  • Dragon_Nexus
    Dragon_Nexus Posts: 3,701 Chairperson of the Boards
    CHRISJN said:
    If like that, hoarding seems no use... it still luck based.

    Wait, were you under some impression it wasn't? All tokens are luck based. I'm not sure what you've been expecting.
  • peanut_gallery
    peanut_gallery Posts: 119 Tile Toppler
    I'd agree with your assessment Piro if every individual token pull was run through the RNG to determine the outcome.  But I don't think that's how D3 works the tokens.  Unless they've completely reworked the RNG process (so that each individual pull is randomized), it's been shown that with each new 5* introduction, each player effectively gets a randomized set of tokens (like a shuffled deck of cards), the order of which won't change until the next 5* goes into tokens.

    I don't know how deep that "deck of cards" goes, but my guess (and I could be wrong) is that the stated probability of pulling a 5* is worked into the distribution of tokens into the deck, but it could take hundreds of pulls, perhaps, to see it. From folks I know who have hoarded huge amounts of tokens and CP, redemptions typically come close to that 15%.
  • Jarvind
    Jarvind Posts: 1,684 Chairperson of the Boards
    edited August 2017
    Unfortunately this is a common complaint and not one that is likely to ever be resolved. 

    Probability is streaky and unpredictable. As @Piro_plock stated, the odds are only "true" when your sample size is "infinity." Each time you pull a 4* from a legendary token, you don't "get closer" to the next 5* pull. You have an 85% chance of not getting a 5* every time you pull. If you could amass thousands and thousands of pulls, you'd probably average out very close to the stated 15%, but even then it wouldn't be guaranteed.

    Only way to reliably cover 5*s is to brute force it with a ton of pulls, and that means hoarding. No way around it. I don't have the post bookmarked but somebody on here did the math and figured out that you need around 300 pulls to reasonably expect to cover someone in a batch of Latests. The numbers on Classics were uh...not encouraging.

    As far as my will to play, I did manage to cover and champ some 5s and I hated the experience so much that I sold them. You can read about it here if you like.
  • peanut_gallery
    peanut_gallery Posts: 119 Tile Toppler
    Somehow I've managed to get 5* covers decently until now (I've got 7 champed 5*), but something seems broken when I've only pulled 2 5* out of what is now the last 76 LL pulls.  Or I just happen to be the unluckiest person in MPQ this season.
  • Bowgentle
    Bowgentle Posts: 7,926 Chairperson of the Boards
    Guess what, I'm 11/90 since Spidey went in.

    I'm pretty sure I'll beat you.

    This is not my first streak like that so yeah.
    At least 3 0/40s.
  • TetsujinOni
    TetsujinOni Posts: 181 Tile Toppler
    So a lesson in how probability distribution works for a fixed-stream seeded random system (which is how you prevent people from one angle of cheating on draws):

    A sufficiently random sequence of values for cryptographic use (that is, knowing the first number doesn't let you predict the second number or third number, etc, unless you know the "secret" piece that chose the particular random sequence you're pulling numbers from).

    This is the "deck of cards" analogy that you cited above Peanut.... the thing you're missing is that the SEQUENCE is probably both player-linked (seeded by the player account ID#, likely) and doesn't change (and it shouldn't.... or you could have a new avenue for multidevice cheating..) 

    So you have this stream of random values with a broad range (billions of them if it's a 32 bit integer function producing the sequence), and a choice between "15% chance" and "85% chance" outcomes. 

    Distributing billions of possible values into 5%/5%/5%/42.5% new/42.5%vintage is what will need to change every time they rotate in new characters, and which characters are in those 5% brackets for the Epics.
  • huktonfonix
    huktonfonix Posts: 214 Tile Toppler
    Should include the last line of that Rod Stewart verse:  Some guys do nothing but complain.

    (Sorry, couldn't resist)
  • peanut_gallery
    peanut_gallery Posts: 119 Tile Toppler
    Lol.  Yes, I do realize I am complaining.
  • zodiac339
    zodiac339 Posts: 1,948 Chairperson of the Boards
    Jarvind said:

    As far as my will to play, I did manage to cover and champ some 5s and I hated the experience so much that I sold them. You can read about it here if you like.
    I thought I recalled that being your post. I saw you recently in my Unstable ISO-8 bracket (right next to another familiar forum name), and saw that lack of 5* Champs. I think you still had a relatively high-level unChamped Phoenix too? Just kind of neat to see those familiar names in-game. (Waves "Hi neighbor!")
  • veny
    veny Posts: 834 Critical Contributor
    Truth is, i am much more bothered by colors and RNG...
    My 0/5/0 Green Goblin or 8/12/2 Prof X says it all :D
  • turbomoose
    turbomoose Posts: 806 Critical Contributor
    I just let my son or daughter open them as they have better luck than me, I'm letting them open 18 LL and about 700 points tomorrow 

    i use my points on classics , as they are cheaper 
  • Alfje17
    Alfje17 Posts: 3,822 Chairperson of the Boards
    edited August 2017

    It's all luck, got 18 5* (+1 bonus) out of 74 pulls, so that's 24% compared to the official 14%. I'm expecting to draw tons of 4* in the future, but there's no telling.


    I posted some statistics on a thread a while back and apart form my LT pulls, it all stays close to the official figures: f.ex. 177 3* (+ 47 bonus) out of 686 Elite tokens, which is 25.8% compared to official 25%.


  • mpqr7
    mpqr7 Posts: 2,642 Chairperson of the Boards
    From what I've read, statistics require that you collect at a minimum 300 pulls (ie 7500 in cp and LT, where 1 LT = 25 cp) to get a new 5* championed in one pull session.

    However for even better odds, you should wait until you have 400 pulls (1000 in cp), 500 (12500) or even 600 (15000) or greater. Even then, it's still not guaranteed, but odds will be higher.

    I started hoarding on May 24, and currently I'm at 70 LT + 2248 cp = 3998 cp = 159 pulls, so it's a very slow and grueling process. Also I rarely hit 1200 in PVP (my work and personal schedule doesn't allow me to make many hops, especially on weekends), nor am I a member of any buy clubs (I just spend $2/month on HP to get access to those sweet, sweet bonus rewards).


    It will probably take me at least another 6 months to get to enough CP to do my pull. I am curious / excited / terrified of the following happening:

    * I still don't end up with a new, fully championed 5*, so it will have been completely useless.

    * Something changes in the way 5*s are created and delivered to players. That will be fine, as long as it's not that I pull my hoard, and then a few weeks or months later, odds are markedly improved. (These things have been known to change! For instance, first it was 10% odds, and then it went up to 15% odds with little or no notice given to anyone).

    * I am successful in championing a brand new 5*, the moment it hits Latest tokens. People start accusing me of being a cheater / hacker. (I'll be okay with this... just as long as CS doesn't decide to sandbox me!)
  • Bowgentle
    Bowgentle Posts: 7,926 Chairperson of the Boards
    11/93 now.
    Well, 10 plus one BH, so 10/93 really.

    I'm pretty sure I'll go 10/104, peanut.
  • Jarvind
    Jarvind Posts: 1,684 Chairperson of the Boards
    zodiac339 said:
    Jarvind said:

    As far as my will to play, I did manage to cover and champ some 5s and I hated the experience so much that I sold them. You can read about it here if you like.
    I thought I recalled that being your post. I saw you recently in my Unstable ISO-8 bracket (right next to another familiar forum name), and saw that lack of 5* Champs. I think you still had a relatively high-level unChamped Phoenix too? Just kind of neat to see those familiar names in-game. (Waves "Hi neighbor!")
    Yep, that's me. :)  I do plan to eventually try the 5* game again, so I kept Phoenix - she was (and is) only at 360, so she doesn't screw with my scaling.
  • atomzed
    atomzed Posts: 1,753 Chairperson of the Boards
    I have been tracking my pulls for sometime. 

    53/ 313, which corresponds to 16.9%. 

    The rates looks good, but if i didn't record it, i would only remember the dry patches. 

    Sometimes when it rain, it pours. i have 1 stretch which i pulled 6 5* out of 10 tokens. That was really lucky. 

    If its any consolation, you may get really lucky and have a purple rain. good luck!
  • OneLastGambit
    OneLastGambit Posts: 1,963 Chairperson of the Boards
    Lots of people complain about 5* pull odds (and it's largely valid but they're so busy focusing on whether or not they could cover and champ ab5* they don't bother asking whether or not they should....until it's too late. 

    Jarvind isn't the only player I've seen despise their transition, snowcatt has also made similar comments in the past too. In particular it's devastating for pvp (pve has been somewhat rectified now) many honest advice is to keep your 5s softcapped to a level where they are about as powerful as your 4s

Categories