Full disclosure please - RE: drop rates

1356

Comments

  • OzzmanOzzman Posts: 1,388 ✭✭✭✭✭
    Fun fact: 1 out of every 100 people will not pull a single legendary from as much as 612 Ceramic packs. :)
    Fingers crossed for @Avenged to not be one of these guys.
  • Blue2moroBlue2moro Posts: 2,004 ✭✭✭✭✭
    m4g said:
    TimBut said:
    Before I quit in last month I pulled Veyron, Ford GT(new one), Pagani Zonda F, Lambo Gallardo, Lambo gallardo, Bmw m5 competition. 

    6 S from about 952 ceramics and maybe 15-20 cfs. 

    But 5 rq 30 S and one 29rq from few ceramics and carbons... 
    1) Dude is lying and he actually opened a **** tone of packs.
    2) Dude has super incredible luck



    Blue2moro said:
    Not only 6 S class in 6/7 ceramics and 3 CF...but those 6 S are the top Bugatti and 5 damn Lambos...all RQ30 except for 1.

    I mean, there is insane pack luck and then there's this.

    My only hope now is that Tim comes along and tells us this account actually spent a metric boatload of cash.

    5 RQ30 & 1 RQ29 from 10 packs...ahahahahahahhaha 😭😭😭😭😭😭😭
    Want a full garage tour?

    If I spent that much I would have more spaces and much better range of other cars.
    Sorry fella, didnt mean that to come across as accusatory against you. Meant more for my own sanity that I wish it wasnt true that someone can pull such godly cars in just 10 packs.

    I suppose for every @Avenged to be an @m4g
  • bantel_catbantel_cat Posts: 936 ✭✭✭✭✭
    0liver77 said:
    Can you provide the maths behind your table? I‘d like to create such a table for my account too. I‘ve notes of all my Ceramics and Carbon openings for more than a year. But I don’t know that to calculate the drop rates. 
    Here you go:
    https://forums.hutchgames.com/discussion/6087/ceramic-opening-analysis
    In this thread you will find the proper Template
    @0liver77 you can get it here: https://drive.google.com/open?id=1qGOsLiT4ELbhP1cE6p_uYiqt0y28fEUK

    I do need to update it.
    The 'Quality' Metrics will be slightly out with the addition of shedloads of cars.
  • bantel_catbantel_cat Posts: 936 ✭✭✭✭✭
    Here is my recent CF Activity:



    Though nothing outstanding actually pulled aside from the RQ30 M5 (and the Alfas that helped me get the Quattro I suppose)
  • AvengedAvenged Posts: 462 ✭✭✭✭
    Cheers @Ozzman, a dude can hope 😉
  • RobGripesRobGripes Posts: 2,257 ✭✭✭✭✭
    Avenged said:
    Cheers @Ozzman, a dude can hope 😉

    You've shown amazing resilience 


    Does anyone have an easy way to keep records on mobile? Android specifically
  • RuvlaRuvla Posts: 71 ✭✭
    Ozzman said:
     I don't think they do, this is how random works.
    You explained how math random works, but computer random doesn't work this way. There is plenty things that can cause deviation when we speak about game, using any kind of random.
  • evilprofesseurevilprofesseur Posts: 1,494 ✭✭✭✭✭
    Ruvla said:
    Ozzman said:
     I don't think they do, this is how random works.
    You explained how math random works, but computer random doesn't work this way. There is plenty things that can cause deviation when we speak about game, using any kind of random.
    computer random is quite enough for this sort of applications.
  • MSteeLMSteeL Posts: 956 ✭✭✭✭✭
    Support no longer retrieves expired cars guys, just to let you know. 
  • OzzmanOzzman Posts: 1,388 ✭✭✭✭✭
    Ruvla said:
    Ozzman said:
     I don't think they do, this is how random works.
    You explained how math random works, but computer random doesn't work this way. There is plenty things that can cause deviation when we speak about game, using any kind of random.
    You're right, real random and computer random is not the same, but in our case they're probably close enouh to treat them as such. I just wanted to explain that it's human nature to expect random events with the same probability to be evenly distributed even with a small sample size. But in reality, strange things happen more often than you'd expect.
  • RuvlaRuvla Posts: 71 ✭✭
    Ozzman, computer random on big numbers will give your exact distribution on long series. Despite to real world where it might not be so, as you clearly explained.

    But, inaccurate implementation can "break" that normal deviation.
  • 0liver770liver77 Posts: 145 ✭✭✭
    @bantel_cat Thanks for the table!
    @mrpiratepete Thanks for the link!
    I created a new table using my data (collected for a little bit more than a year):

  • Hutch_TimHutch_Tim Posts: 627 admin
    Okay, we've said it before, and it remains true: all drops are completely independent of each other and random. There is no bias to new or old players, or anything like that.

    As I noted a long time ago though:
    https://forums.hutchgames.com/discussion/comment/52637/#Comment_52637
    ... if you don't trust us on the drop rates, then obviously you won't trust us when we say they are random, so carefully controlled player trials are the only way you can be sure of that.

    Now, analysing the least lucky and most lucky people isn't really the best way to analyse a population probability problem, but two examples in this thread did stick out.

    First, @Avenged does look unlucky, but also not very active? Assuming I looked up the right account, it looks like you opened an average of 1.6 Ceramics a week over the last year. Some people here open the same amount in a single month, so if that's really the case I'm afraid it's less surprising your dry spell has been so long.

    Then there's this @m4g example. Now, we do have quite a lot of new players coming into the game. Imagine 10,000 versions of yourself all starting Top Drives in parallel universes, and then imagine the luckiest one of those 10,000 versions. Well, we'll have several players that lucky every week.

    Even so, the stated packs and drops look weird. Looking up the account, they have opened more packs than they said - I make it 19 Ceramics, 3 Italian premiums, 1 IR premium, and miscellaneous others including some CFs - but even then, they would probably be one of the luckiest players in the history of Top Drives. So that's quite... interesting.
  • hillclimberhillclimber Posts: 2,156 ✭✭✭✭✭
    Ozzman said:
    this is a topic where bias kicks in, I have spoken out in that regard before. but after spending 5+ millions on IR Premiums and getting nothing more than 8 URs mostly Maseratis I'm can't fight my doubts any longer. of course it is possible to happen, as much as it possible to win the lottery, just the other way around. I only wonder what the odds are. could someone here help me to do the math?

    5 millions on IR Premiums equals about 120 packs. what are the odds of getting 8 URs and nothing above with these drop rates? 

    Let's simplify a bit by concentrating on Card 1 only. The chance of not pulling an A or S there (and let's assume that means the whole pack) is 97.55%. So if you open 120 packs, that means 0.9755^120=0.051, meaning around 5% chance of not getting anything above B even after 120 packs. Unlucky, but happens to 5 out 100 people. Card 2 changes this a bit, but just slightly (my guess would be around 4.5%).
    thanks a lot for helping me out here. I agree these odds are in the range of possible. I'm feeling better now 🤥
  • AndreasSimmerAndreasSimmer Posts: 1,046 ✭✭✭✭✭
    What about a boost for low rq players, or packdrops boosted relating to coming/running events?
    Especially the second thing interests me, as i know its not just me who could predict coming events pretty often just by anaylyzing the drops.
    Its not that every car relates, but it feels these get a boost, as do new update cars seem to have aswell.
    About @m4g, lets say he opened 30 packs, and really got more Rq30 cars than i did in 2 years intense playing. What are the chances that all of these happen to be italian cars if there is no such booster?
  • Blue2moroBlue2moro Posts: 2,004 ✭✭✭✭✭
    What about a boost for low rq players, or packdrops boosted relating to coming/running events?
    Especially the second thing interests me, as i know its not just me who could predict coming events pretty often just by anaylyzing the drops.
    Its not that every car relates, but it feels these get a boost, as do new update cars seem to have aswell.
    About @m4g, lets say he opened 30 packs, and really got more Rq30 cars than i did in 2 years intense playing. What are the chances that all of these happen to be italian cars if there is no such booster?
    Probably better odds of winning the lottery and buying your own Lambo!
  • Hutch_TimHutch_Tim Posts: 627 admin
    Oh yes, to elaborated further: there's also no difference in odds for low RQ players, and no bias of drops towards upcoming events. As I think I also said last time, if we really did bias drops towards upcoming events we would shout about it as it would presumably boost pack sales! 
  • O__VERO__VER Posts: 1,569 ✭✭✭✭✭
    Hutch_Tim said:
    Oh yes, to elaborated further: there's also no difference in odds for low RQ players, and no bias of drops towards upcoming events. As I think I also said last time, if we really did bias drops towards upcoming events we would shout about it as it would presumably boost pack sales! 
    I have to ask, @Hutch_Tim, because there is just SO MUCH of what feels like good evidence to the contrary — how do you know this? Is there any auditing done on your developers' code? How do you know that your developers haven't added in this type of thing without anyone else's knowledge?
  • baestbaest Posts: 507 ✭✭✭✭
    Hutch_Tim said:
    Oh yes, to elaborated further: there's also no difference in odds for low RQ players, and no bias of drops towards upcoming events. As I think I also said last time, if we really did bias drops towards upcoming events we would shout about it as it would presumably boost pack sales! 
    Are said algorithms audited by any 3rd parties?
  • RobGripesRobGripes Posts: 2,257 ✭✭✭✭✭
    edited October 23
    baest said:
    Hutch_Tim said:
    Oh yes, to elaborated further: there's also no difference in odds for low RQ players, and no bias of drops towards upcoming events. As I think I also said last time, if we really did bias drops towards upcoming events we would shout about it as it would presumably boost pack sales! 
    Are said algorithms audited by any 3rd parties?

    This is the main question to Hutch.

    Undoubtedly we understand odds enough to know someone could be extremely lucky. I remember a player who got banned (unrelated) last year who had two, yes two double legendary pulls from ceramics...

    What I would like to know is has the code been audited? 
  • TGPDTGPD Posts: 5,596 ✭✭✭✭✭
    While I don't think a reparation scheme is any good or is necessary, I think we can all agree that


    The drop rates should be improved across the board.
  • GsearchGsearch Posts: 2,475 admin
    TGPD said:
    While I don't think a reparation scheme is any good or is necessary, I think we can all agree that
    The drop rates should be improved across the board.
    I can't agree with this. the fact is that it would not change a thing.  If Hutch increased the drop rates an all packs the outcome would still be the same. 
  • RobGripesRobGripes Posts: 2,257 ✭✭✭✭✭
    Gsearch said:
    TGPD said:
    While I don't think a reparation scheme is any good or is necessary, I think we can all agree that
    The drop rates should be improved across the board.
    I can't agree with this. the fact is that it would not change a thing.  If Hutch increased the drop rates an all packs the outcome would still be the same. 

    How do you mean Gsearch, can you elaborate which outcome would remain the same?

    I think the run of bad luck would be shortened in the extreme cases and the luckiest of players can still only field 5 cars. I don't think anything drastic needs to change, but if we say someone opens 1 ceramic a day and one CF a week, they could only expect a small number of legendaries a year
  • RobGripesRobGripes Posts: 2,257 ✭✭✭✭✭
    Gsearch said:
    @RobGripes  What I mean is that if the change was for every player then the outcome would be the same. If the drop rates went up 10% then EVERY player would have a garage that was 10%better than it is now and the event results would be the same.

    Yeh i agree largely with this. Only one point to supplement though is that there is a ceiling with which one could play their hand. Five maxed cars. 

    At the end of it, i do agree with TGPD that the drop rates should be improved a bit.
  • RWareRWare Posts: 904 ✭✭✭✭✭
    Whilst I don't think different accounts roll different algorithm's, those results from that amount of packs is sure hard to swallow.

    Not shouting foul play, it's just the amount of salt in my throat...
  • mario_mmario_m Posts: 100 ✭✭
    edited October 23
    Ciao
    beyond my feeling that the drop of new players is more fortunate, it is a feeling not confirmed by data so I am certainly wrong.
     @Hutch_Tim  is it possible that for those who play on 2 devices there are problems on the drop?
Sign In or Register to comment.