Jdberia wrote: I forked over the HP to by a 10-pack when Psylocke was the guaranteed pull and got ONE. And the rest of the covers were mostly for characters I don't use. I know what we can pull is all random but I was not happy.
allorin wrote: Jdberia wrote: I forked over the HP to by a 10-pack when Psylocke was the guaranteed pull and got ONE. And the rest of the covers were mostly for characters I don't use. I know what we can pull is all random but I was not happy. This. I stupidly (last time ever) forked over the money for the HP to buy a ten pack after being frustrated at being locked out of Hulk nodes. I got one Blue BP, 2 Thors and SEVEN Modern Hawkeye covers. I was honestly £&@#ing disgusted at that. There are no 2* covers I need, but to throw 7 M Hawkeye covers at someone is taking the piss. I resolved then and there that MPQ isn't getting another penny from me.
HailMary wrote: allorin wrote: Jdberia wrote: I forked over the HP to by a 10-pack when Psylocke was the guaranteed pull and got ONE. And the rest of the covers were mostly for characters I don't use. I know what we can pull is all random but I was not happy. This. I stupidly (last time ever) forked over the money for the HP to buy a ten pack after being frustrated at being locked out of Hulk nodes. I got one Blue BP, 2 Thors and SEVEN Modern Hawkeye covers. I was honestly £&@#ing disgusted at that. There are no 2* covers I need, but to throw 7 M Hawkeye covers at someone is taking the piss. I resolved then and there that MPQ isn't getting another penny from me. Ouch. I'm really curious as to the probability breakdown for specific packs during specific times. M.Hawkeye, specifically, seems to be given out like candy in current Heroics, much as Cap seemed to be during Brotherhood. Perhaps they tweak cover-pull probabilities from event to event, in order to increase the chances of specifically pulling the worst buffed character. Or, maybe it's just a lot of reporting bias. I dunno.
Lyrian wrote: Probability is the same as it always is for a heroic pull - 1 in 6. That being said, event tokens exhibit a very narrow window of obtainable characters, unless stated otherwise on the token description. For this event, the Hulk Comic Pack generates "Powered-Up" heroes, which limits the character possibilities for 2**s to just three: Modern Hawkeye, Thor, and Original Black Widow. As one of these three characters has an approximately 83% chance of being pulled, many copies of a single character are not uncommon. For none of the other 9 characters to be any of the 3*** characters, the probability is (5/6)^9 = 19.38% Thus, there is a 1 in 5.16 chance in every 10-pack that none of the pulls beyond the guaranteed one will contain any 3***s at all. Tough odds to chew on, when essentially buying a $30 lottery ticket for 3800-4000 hero points.
HailMary wrote: Lyrian wrote: Probability is the same as it always is for a heroic pull - 1 in 6. That being said, event tokens exhibit a very narrow window of obtainable characters, unless stated otherwise on the token description. For this event, the Hulk Comic Pack generates "Powered-Up" heroes, which limits the character possibilities for 2**s to just three: Modern Hawkeye, Thor, and Original Black Widow. As one of these three characters has an approximately 83% chance of being pulled, many copies of a single character are not uncommon. For none of the other 9 characters to be any of the 3*** characters, the probability is (5/6)^9 = 19.38% Thus, there is a 1 in 5.16 chance in every 10-pack that none of the pulls beyond the guaranteed one will contain any 3***s at all. Tough odds to chew on, when essentially buying a $30 lottery ticket for 3800-4000 hero points. Fair enough. I suppose it's reporting bias, then: there seems to be a lot of people getting a preponderance of MHawkeyes from Hulk tokens. It doesn't help my subconscious perceptions that I've pulled 3 MHawkeyes out of 5 Hulk tokens (which admittedly is statistically meaningless).
gobstopper wrote: The google doc mitigates some of the bias: https://docs.google.com/spreadsheet/ccc ... REE#gid=45
Lyrian wrote: Completely agree with your statements, Jachdo. However, in this situation, the OP paid the 3800 HP because he/she didn't have any Black Panther covers and wanted the guaranteed cover through the 10-pack. Heroic tokens can't pull a Black Panther at the moment.
Jachdo wrote: That's the probability if a truly random draw were occurring, which it almost certainly isn't DISCLAIMER: I'm not a math guy, so please try to contain any nerd rage and accept my apologies in advance! I'm gonna start by saying that computers themselves are unable to generate truly random numbers. There are many advanced algorithms out there that use a seed to generate what appears to be a random integer, but really it's just part of a larger sequence we can't perceive. Unless Demiurge has Skynet running their servers, the computer has no free will to choose. The only other alternative would be a correlate the number with naturally occurring random event. A well cited example involves hooking a computer up to a Geiger counter which is monitoring a piece of Uranium. Since radioactive decay is a completely random naturally occurring event, the data received from the counter allows the computer to make an uniformed, and truly random number. Unfortunately (fortunately?) it's also pretty unlikely that there is a piece of weapons-grade plutonium in the Demiurge server farm. ...
Phantron wrote: Ug, this is a personal pet peeve of mine, but the whole 'random is not really random' is totally meaningless in the context of a game. Let's say that the cover generated is simply a sequence that still satisfies the fact that all 2* are equally rare, and 3* appears 1/6. If you buy the 10-pack they give you a random Black Panther first and then just read the next 9 covers on the sequence, while buying 1 at a time it just gives you the next cover in the sequence. Such a sequence might look something like: Hawkeye, BP, Thor, Thor, OBW, Punisher, Hulk, Hawkeye, Thor, Hawkeye, etc., etc. Let's say this sequence is only 1000 covers long (even though it's trivial to generate a much longer sequence) and you were handed the list, and you noticed that Black Panther appears on cover #16 and cover #10-#15 are all Thors and this is the only sequence of 6 Thors in a row in the overall sequence. So you know if you get 6 Thors in a row the next one is Black Panther. What good does this do for you? Absolutely nothing because you have no idea where on this sequence you're at, as this sequence is handed out to all the players of MPQ. On average you're nowhere near the '6 Thors in a row' position. Even if say a guy just posted like "I got 6 Thors in a row this is messed up!" (and this guy doens't know about the trick), as long as there is more than you that knows this secret, there's no guaranteed you'll beat any other guy who just read this. For that matter, someone may have bought that cover #16 before the guy with 6 Thors even let you know of his misfortune. This is a system that isn't random at all, and yet the outcome of this system, to any observer, is effectively random because you simply have no way of knowing what position on the sequence you're in. The whole 'pseudorandom doesn't approximate reality' is because no one is sure if the decay of an Uranium atom can be captured by a formula. We know the half life of Uranium is X years but that's an observation. We don't know what exactly causes the atoms to decay at that rate so we're not sure if you can simply just say '1 atom decays every Y seconds' even if that's what math predicts. But there's nothing you are trying to approximate in the cover distribution. Getting 1/6 cover being 3* is a defined criteria, not an observation.
Jachdo wrote: DISCLAIMER: I'm not a math guy, so please try to contain any nerd rage and accept my apologies in advance!