Location Coney IslandPosts: 259 Amazing

We got into a discussion buried within another thread about the "Gambler's Fallacy", but it is an interesting enough topic to warrant its own thread as it may come up in your life at some point. Scenario: 1) I hold a fair coin in my hand and ask you, "what are the mathematical odds that I flip 5 heads in a row?" 2) You do the math and say, "a 1/32 or about a 3% chance". and, you would be correct. 3) I proceed to flip 4 heads in a row, then stop and ask you, "what are the odds that this last flip is a heads again?" Here is where a mistake can be made if you don't keep your eye on the ball. If your answer is "1/2 or 5050" then you are correct and did not fall for the "Gambler's Fallacy" as it is known. However, if you thought to yourself, "well, it is the same conditions as it was before; it is still trying to get 5 heads in a row which is a 1 in 32 chance." And your answer was "3%", then you got Jedi mind tricked. (*thanks to Popinski for idea on rewording this section) So how did you answer? Here is the issue of Gambler's Fallacy explained on Wikipedia: en.wikipedia.org/wiki/Gambler%27s_fallacy 
Location beyond the Black HolePosts: 1,282 Fantastic

RNG is fake anyway... ever tried to play the same 400 mp3s in winamp in shuffle mode with the random song button on? after listening to the same list in the same mode over and over again, starting with the same song in the list, the order which should be "random" turns out to only play the same few songs over and over again, so you look through the list and see a bunch of Songs you haven't heard in weeks or months and some that keep running like every other hour. what if i told you, there is no RNG? it is all a pre defined pattern, your sample size is just o small to see it. but i have yet to witness Winamp playing the same song 4 times in a loop while set to shuffle. hands out tinfoil hats 
True BelieverPosts: 247 Amazing

The gamblers fallacy isn't someone saying you have a 3% chance to flip the coin on the 5th time. gamblers fallacy is seeing patterns when there is none. The people who think they found that lucky slot machine. Or they swear by there being a higher chance of getting uniques if they swap to their level 1 before they start a terminal. Your example has the "gambler" not noticing a pattern which is the exact opposite of what the fallacy is. Of course in this game it is hard to tell between an actual gamblers fallacy or a possible bug. For all we know there might be some big bug effecting drops and RNG. 
Posts: 62 Mighty

There are different kinds of randomness. In games they usually use a more evenly distributed "randomness" formula, so it will keep players from complaining about getting the same thing twice in a row and it should be like that, else it would be chaos (even so, sometimes it happens).
Winamp even has a slider for "randomness" so it's obvious it has a pattern to not play the same songs over and over again, except if you rate them, then it will have a higher chance of playing the songs that you like more. It's actually a formula for that. I thought these things were obvious to begin with... Even if you flip a coin, the side it ends up on is governed by the laws of physics. 
Posts: 2,821 Fantastic

I'd say the 5th flip has a 100% chance of landing on heads, then shoot you for being a damn cheat. Cowboy laws. 
Location The White Hot RoomPosts: 375 Amazing

Computers are fundamentally incapable (as are humans, as a note) of generating truly random numbers. When you see things like Winamp's "randomness" slider, what you're actually seeing are features that implement more complex (read: less efficient and more memory hungry) mathematics to generate numbers with a higher degree of unpredictability, usually by cycling through a multilayered set of random functions where: In the case of music players, additional weighting is often done based on rating systems, or "Likes," if you're fond of Pandora or whatever the flavoroftheweek internet radio service is. This is usually done by weighting the possible results of the randomly selected number rather than the number itself though  it's much less computationally demanding to randomly select from a pool of 100 songs you 'Like' and 25 you don't than it is to weight the numbers towards the songs you like in a list of hundreds (of thousands) of entries. 
Posts: 1,684 Fantastic

To be fair, technically, nothing is random due to cause and effect. The only issue is that some things are effectively random because the cause cannot be adequately ascertained. 
Location CaliforniaPosts: 2,252 Fantastic

You worded that completely wrong, but yeah, that's sort of right. 
Posts: 149 Incredible

Agree. Cause and effect=everything. That's what Lambert said to Keanu (and he didn't understand). 
Location The Penalty BoxPosts: 842 Amazing

You are completely wrong. Gambler's Fallacy has nothing to do with superstition. It's a misunderstanding of probability. It's thinking that a coin is "due" to flip tails because it just flipped heads 4 times in a row. It's thinking you "used up" your good rolls because a unique dropped on your level 5 toon. 
Location Coney IslandPosts: 259 Amazing

The wiki page has everything on the subject of course, but here is a relevant quote: "While a run of five heads is only 1⁄32 = 0.03125, it is only that before the coin is first tossed. After the first four tosses the results are no longer unknown, so their probabilities are 1. Reasoning that it is more likely that the next toss will be a tail than a head due to the past tosses, that a run of luck in the past somehow influences the odds in the future, is the fallacy" en.wikipedia.org/wiki/Gambler%27s_fallacy I got it right. So I don't know where the "completely wrong" you wrote is coming from. All I did was put the same information from the wiki page inside a conversational form to make it easy to follow. 
Location CaliforniaPosts: 2,252 Fantastic

You worded the conversation wrong. "What are the mathematical odds that I flip 5 heads in a row?" "A 1/32 or about a 3% chance" I proceed to flip 4 heads in a row, then stop and ask you, "what are the odds that this last flip is a heads again?" "Its the same as it was before, its still only a 3% chance to do 5 in a row." The guy didn't really answer your second question, but he gave you a true statement. Just a grammatical inconstancy that can confuse the reader. 
Location Crazy Town, Banana Pants Posts: 2,482 Cosmic

I am not saying you are wrong, but Wikipedia can be edited by anyone. While it does have a large amount of correct information and a lot of people put effort into making sure it is correct and that it stays that way. It also has an alarming amount of incorrect information. So it is not always an airtight reference source. Love the thread, interesting read. just thought I would throw that out there. 
Location Coney IslandPosts: 259 Amazing

Please go read the wikipedia page about "Gambler's fa> @Scorpiodisc said:
Thanks. And of course I agree with you. Anyone who as done at least one research paper in the past few years knows that Wik will not fly as a source. But I felt for such a lightweight subject and for easy reading I would go with the Wik. If I had to get serious about it, I would site this instead: mansci.journal.informs.org/content/39/12/1521.short What is creepy though, is even when wording my example like you would talk to a child to make it simple, and giving a reference, some will come back and post that I have no idea what I am talking about. 
True BelieverPosts: 247 Amazing


AssembledLocation Some place nice! :)Posts: 944 Amazing

Statistics is really not that hard. It's just that people are often counting the wrong sample space 90% of the time. 
Posts: 409 Amazing

I tend to suffer more from confirmation bias  if it works (0% RIF and 56% SIF) and generates excellent results over time (4 uniques, >10 cosmic, >30 artifacts) I'll keep doing it. I did level up colossus with pure +RIF, but my opinion is there's a bug as I tend to get better rewards with 0% RIF. 
Location Coney IslandPosts: 259 Amazing
