I don't pay for Keys because I have alts' drops to get me free Keys.

However, I would be better off if I just saved up for items... eh, have just about everything I'll want ingame (except for 'nother Unusual, har) regardless.

I think I need to stop for good if I somehow win big.

Playing Flyff, I was trying to train one of my guild lieutenants on making money. I had bankrolled him and then I hear "Well I'm making money on this dice game!" I figure he has to learn his lesson...the next day I log in and all the guild money is gone, and so is he. Hope he never goes to Vegas.

@Kaja Rainbow: I didn't mean my point to make a rebuttal. I was more meaning how frustrating these losing streaks can be, particularly when the odds are in your favor.

Fucking illiterate dumbass kids. Are you fucking 12 or did you even finish high school? No wonder the welfare rate is so low, people can't even count to ten, and they spread misinformation like nutella on their noname(c) white poor man bread.

E.G, when you flip a coin, you can expect 1/2 tails, even if you yourself desire a head. What you expect =/= what is expected objectively.

Look at the definition of expected value: http://en.wikipedia.org/wiki/Expected_value

[url=#user_comment_138255]@Sircalist[/url]: No, you are retarded and you don't understand the law of Large Numbers. EXPECTED VALUE DOES NOT REFER TO WHAT YOU EXPECT, it refers to the median average most probable possibily to be expected objectively.

MADE A MISTAKE IN THE PREVIOUS COMMENT - in the picture, it should say DECREASE, not increase in the top right corner. I apologize for this typo, even I am not perfect!

CORRECT IMAGE: **** http://i48.tinypic.com/346b968.jpg ****

The image in the comment below has a typo.

Look at this picture: http://i45.tinypic.com/fbb7ro.jpg I made an actual probability tree from opening boxes. You can see that your odds to get an unusual DECREASE each pick. Also, here is the program I used: http://kera.name/treediag/ If you want I'll upload the file and you can check my diagram.

Actually the Law of Large Numbers states that the average of the results of an experiment should be close to the expected value. e.g The more you unbox "unsuccessfully", the more probable is to unbox a rare item. However, it's statistics, you can't be sure to unbox. Kids, never spend money on keys.

HAH, suck on that, Fate.
Just unboxed an unusual.

And I almost forgot, this is NOT ON THE SHEET:

In percentage odds of 225/4096 from

Lose win lose
Win Lose lose
Lose lose win

All those that are 2 lose and 1 win

is 5,5%

The percentage chance to get a single win is 6,25%

If you reply I will reply if not goodbye. It's stimulating, waiting for DIII

And it will be even lower.

Now, you may not care anymore. This was an interesting, challenging discussion. Hopefully you learned something; I admit statistics are kinda hard- I struggled to explain I must admit, but I'm a self called math teacher.

http://i45.tinypic.com/vo95dj.jpg

Here. Everything's explained on that picture. I've even included the formula, and the problem is based on the 16 sided dice problem from wikipedia, problem in which you claim your odds to win increased with every roll. I went up to three. Go up to 16...

Got another thing coming, another tree this time with the dice experiments. Wait about 15 mins.

If you look at the end, the odds of picking a red are lower than at the beginning, sometimes even, depending on the serie.

Bonk!

http://i46.tinypic.com/2ch4bw9.jpg

On the first tier, you have 5/10 chance to get a blue, 4/10 of a green, 1/10 of a red. On the second tier, it's the same. However, your odds of getting a serie can be calculated.

I'm working on something now and should be done in 20 minutes or so. I'll show you why odds doesn't increase the more you play/do/open.

Be back in 20 mins.

64.39% is still higher than the chance of winning one roll, which is 6.25%.

It all depends on your point of reference, from the beginning of the set(Increasing per chance), middle of the set(Decreasing per roll), or end of the set (Same for any one roll)

The player becomes more likely to lose in a set number of iterations as he fails to win, and eventually his probability of winning will again equal the probability of winning a single toss, when only one toss is left: 6.25% in this instance.

a series of losses; his odds have decreased because he has fewer iterations left to win. In other words, the previous losses in no way contribute to the odds of the remaining attempts, but there are fewer remaining attempts to gain a win, which results in a lower probability of obtaining it.

Simply by losing one toss the player's probability of winning dropped by 2%. By the time this reaches 5 losses (11 rolls left), his probability of winning on one of the remaining rolls will have dropped to ~50%. The player's odds for at least one win in those 16 rolls has not increased given

However, assume now that the first roll was a loss (93.75% chance of that, 15⁄16). The player now only has 15 rolls left and, according to the fallacy, should have a higher chance of winning since one loss has occurred. His chances of having at least one win are now: 62,03%

You misinterprated wikipedia.

The low winning odds are just to make the change in probability more noticeable. The probability of having at least one win in the 16 rolls is:

64,39%

We're not talking about statistical dependency or the odds of a SINGLE chance at this point. We're talking about probability of success in a SET of chances. The odds of getting AT LEAST ONE success increases with the size of the set, even though the success of any one toss is the same.

in the following way: what makes a sequence random is that its members are statistically independent of each other. For instance, a list of random numbers is such that one cannot predict better than chance any member of the list based upon a knowledge of the other list members.

Both fallacies are based on the same mistake, namely, a failure to understand statistical independence. Two events are statistically independent when the occurrence of one has no statistical effect upon the occurrence of the other. Statistical independence is connected to the notion of randomness..

Scow2:

No. This is called Gambler's fallacy. The only way you can increase your odds is when you don't put the item back in after the pick, reducing the pool from which you pick.

Your odds are based on: When opening a box, you have 1 chance out of 100 to get a rare.

In the example on Wikipedia you linked: Getting a win out of 16 chances with a 1-in-16 chance is actually only 64.39%. However, Getting a Win with a single roll out of a 1-in-16 chance is only 6.25%.

So, the more boxes you open, the higher the chance of AT LEAST ONE of those boxes is good.

The reality is that in one single box, you have 1/100 chance to get a rare. Knowing this, opening 100 boxes would not change the odds to get a rare.

Just like playing and playing and playing at the casino doesn't increase the odds of winning (at random games, like roulette, slots...)

[url=#user_comment_137203]@Math[/url] Teacher&Student. You're both dumb. The more chances you have in a set, the higher the probability of getting a positive result. As your chances remaining approaches one, your probability shrinks down to the one-shot success chance. AT NO POINT is the probability lower than this.

Your theory would not even work even be if the reality would be that 1 out of 100 boxes would give you a rare. Sure, if one 1/100 box would give you a rare, then 2 out of 200. 3 out of 300... but... if you reduce your fractions to the most lowest denominator... it's still 1/100.

At a lesser scale. A box with 10 beads. 1 red, the rest blue, green, pink.

You have 1/10 chance to pick the red.

You have X picks. The first pick is 1/10. Put the bead back in.

The second? 1/10 as well.

The problem is that 1/100 is not 1 box out of one hundred will give you a rare, it's that from ONE box you have one chance out of a hundreds to get a rare.

Then same for the next one. The odds stay the same.

Look.

I mean because it's multiplied. I really should learn to check before posting...

But the dice example doesn't really fit, because we don't have a set number of tries. For each individual try, the chance is always the same, but the more you try, the lower your chance of failing is multiplied. Thus, your chance to fail 500 times IN A ROW is 0.6 percent.

The example here proves it:

http://en.wikipedia.org/wiki/Gambler%27s_fallacy#Other_examples

The more you roll the dice, the lower your chance to win is.

LEARN TO FUCKING MATH FFS *nerdrages like a mfer on meth*

But, opening several crates increases your chances of getting good items.

You claim to not have failed at math then you say this.

Go read: http://en.wikipedia.org/wiki/Gambler%27s_fallacy

Your chance of getting a good item is always 1% no matter if it's the 1st or 1000th crate.

It's no use explaining really, if they failed at math they will not magically understand it now.
The chance per crate are always the same, ALWAYS. But, opening several crates increases your chances of getting good items. However, the chance per specific crate is still the same.

got an unusual on my first uncrate.
luck doesn't get better than that.

Please go back and finish high school. It's not the 1%, it's you who are retarded and that is why you are poor.

And they all feel the same, look the same, until you take it out of the box and the box vanishes. Then you move on to the next box, and still choose one out of 100 hats, with 99 worthless hats.

That is how it works. Your pool of good/worthless hats doesn't change.

So you pick a hat, and you have 1/100 chance to pick the unusual good hat and 99/100 chance to pick a bad hat.

Then you move on to the second box. But the odds are unchanged. Imagine it in your mind. You're putting your hand in a crate, choosing one hat out of 100, every time.

Example: You have 20 CRATES with 50 worthless hats, 30 bad hats, 19 crappy hats 1 unusual rare good hat inside of each crate. 100 hats total. You can pick ONE hat per crate ONLY, then you need to pick from another crate with the same amounts of hats.

[url=#user_comment_137447]@V[/url]: Fucking retard. Probabilities DO NOT WORK LIKE THAT in that case. Did you finish high school or are you twelve trying to sound smart? The odds of an unusual in that crate DO NOT change the odds of getting an unusual in the next crate, ever. Each pick is irrelevant to the next.

It's like a Random Number Generator. It's possible to get 100 7's in a row but if it keeps doing it, it's not random anymore is it?

Ding ding ding! asd has got it!
If the chances of an unusual is a definite and fixed 1%, then opening up 500 or so is almost guaranteeing you at least one unusual.

If i remember my stats correctly, it's not that every try increases the odds of success, but rather every try has the same odds, so 99 consecutive crappy boxes must mean a very low success rate. Odds are the 100th box is also crap.
P.S. Just cause it CAN happen doesn't mean it WILL.

These chances dont dissapear if you look at every try. so now lets see the chances of the 100 being a bad one too.
The Chance of getting a "bad" crate after 99 bad crates ist 36,6%. Ofc that doesnt mean, that you will get a good crate however the chance of 100 "bad" crates is 36,6%.

ok we open 100 crates
each crate has a chance of 1% to get a thing you want.
the chance of not getting the thing you wanted is 99%.
if you look at each crate the chance does not change.
You got 99 "bad" crates and each crate you had a chance of 1% to get a "good" crate.

@Jake'm: I think what he's confusing this with is the fact that openning 10,000 crates at once would multiply the chance of getting what you want by 10,000, 'course that also multiplies the chance of not getting it by 10,000, but it sounds better when you don't say that.

[url=#user_comment_137215]@anon[/url]: Bull. If there's a finite amount of chocolate, the chance of getting a golden ticket increases with each choccie taken. The more people without a ticket means your chance is higher.

[url=#user_comment_137447]@V[/url]: No, no, no. 10,000 crates could be opened, and no matter what's in them, opening one won't have an affect on the next one. They're independent of each other.

I think the biggest obstacle to understanding statistics is understanding that statistics change as relevant info becomes available, and knowing what is relevant info.

What you got from the last 40 crates is not relevant info, but knowing which of the two doors you didn't pick hides a goat is.

Yes I referenced it, although this might be better called Schrodinger's Crate instead.

Probability is a bitch and makes my head hurt.

Wait are you discussing Jose post or the Scout dialog?
The more you do something the better the chances of getting what you want. It may take 1,000 crates but his chances increase with each crate opened.
Can't think...any...mor..
/headexplode

your mother does meth

and i mean meth

The contents of a crate do not affect the contents of the next crate. You could open a thousand crates and get Strange Diamondbacks. You could open three crates and get three Unusuals. It's dice-rolling, not grabbing balls from a bag.

Gonna say, I see this happen, get suggested to "unbox Unusuals to get money" and every time I die a little inside trying to explain statistics.

But it's not the case, the probabilities shouldn't change each pick in TF2 crate case.

Unless what fluffyneko-san said is true, in which case odds may change.

[url=#user_comment_137215]@anon[/url]: Actually it does if you're going down the line. If the number of elements changes each pick, then the probabilities will change as well....

TECHNICALLY, that is true Jo. However, enough people bitch about it that many games now filter their random number generations so that people don't feel as cheated.

key word: FEEL

It is how it works in TF2: Basically, anything you got, is put back in the "possible drops" pool inside the crate. Odds are then unchanged.

A variation of the experiment is not putting the beads back in, increasing your chance to get a bead of another color.

Additionnally, an experiment to test odds is this simple one:

You have a jar with 4 red beads, 3 yellow ones, 2 greens and 1 blue. You pick three beads one after the other. After you pick one, you put it back in the jar. The odds for the next pick are unchanged.

Nothing to do with Poisson distribution. (http://en.wikipedia.org/wiki/Poisson_distribution) In fact a simple tree to view odds would be sufficient to figure this out.

[url=#user_comment_137396]@Gualdhar[/url]: Yes and no. What Yoyogre is talking about is, the more crates you open, the higher % chance to get something good increases. It doesn't work because the % chance of the current crate you open doesn't affect the % chance of the next crate, and so on for every new crate.

@Platypus

The ME packs are a bit better than crates as once you max out a weapon/upgrade, it stops appearing, giving you a better chance at what you haven't got.

With crates, it doesn't matter in the slightest what you've gotten before.

This also applies to the Mass Effect 3 Multiplayer and their Packs. Actually even more so because you actually can open up a Pack and get virtually nothing of value (as you can get repeat character cards after you've already gotten for class maxed out in appearance options and levels).

Phil, did you just quote Schrodinger's Cat?

Isn't it only gambling if you have a chance of putting something in (ie: money) and getting nothing out? After all with crates you technically can't lose, you just might not get something rare. Plus there's no way to sell your items for money without going against the EULA.

Wow, my statistics class is coming in handy!

Yoyogre you're comparing a poisson distribution (the likelihood a very rare event occurs in n trials) to a single binary event. The chance a single binary event is successful (i.e. the chance a single crate has something good in it) does not change.

The crate contains something that is both nice and crap. As long as you don't open it, the contents will remain both, only upon observation of the contents does it become one or the other. So the lesson is to collect crates and never open them.

Technically, if you open crates and you stop when you had opened 100 or you get "that one drop", you have better chance to succeed than if you stop at less crates.
tl;dr The more you open, the more it is likely.

Reminds me of Fallout games and the 95% chance to hit. My character could fire 3-5 shots per round and in many cases I did not manage to hit anything.

I dunno what the problem is, almost every time I've used a key I've gotten a strange, and the one time I didn't I got a hat... Am I just really stupid and not noticing the vales of the items? :/

Traded a couple of hats I didn't want for a crate because I wanted to play the hat lottery. I chose a crate that had a lot of good stuff that I'd use. In fact, only one weapon was something I didn't want.

Unboxing loot. 3.. 2... 1...
You have unboxed Strange Big Earner...

Story of my life.

I hope people are trolling in the comments and aren't really serious... but probably they aren't trolling this time...

I remember that the old Civilisation games had a buggy probability calculator that would report like ninety percent chance of victory on attacks that would in reality only have about one percent chance of succeeding.

Possibly relevant?

[url=#user_comment_137309]@Malygon[/url] It's Confirmation Bias, I believe.

There was only a 19% chance scout could get at least 1 unusual out of 21 crates. That said, the last crate still only has a 1% chance of success.

oh boy, will people ever learn the difference between statistics and stochastics?

[url=#user_comment_137317]@Wolfzoon[/url] Looking before anyone draws, yes, everyone has 1/10 chance. Once one fails, though, all remaining now have a 1/9 chance. To make it easier, skip to the last person. If everyone else failed, they have a 100% chance. I doubt it works like that, though, so it's pretty much moot.

Still[url=#user_comment_137263] @wkz[/url]: Let's analyze that, shall we? Lets say we have 10 eggs. The first person has 1/10 chance to get it. IF he fails, so that's 9/10 chance, then the second person has a 1/9 chance to get it. 1/9 * 9/10 = 1/10 As he said, the turn order does not matter.

[url=#user_comment_137263]@wkz[/url]: Apparently, you fell for the fallacy while trying to be smartass. How tragic. Let's quote the statement: "Probability-wise, it doesn't matter when you get your turn." And this is what you state: 'if the people before you didn't have the golden ticket you have a reduced chance'

There is only a 37% chance you won't get at least one unusual if you open 100 crates. If you open 470 crates, there is a 99% chance you get at least 1 unusual.

Then you step back and realise you spent $250 and $1175, respectively.

But the 90% of times he predicts the weather correctly you'll not even notice it and just take it for granted. If wonder if that specific problem has a name...

But when something bad happened I noticed it and thought it was deliberately against me. It's a problem with perception. When a weatherman predicts that the sun will shine the next day and you go outside without an umbrella and it rains you'll be angry at him.

Of course he did fall. Not even that, falling damage was determined and he died. And those things happened so often it wasn't funny. But in the end I guess it just comes down to perception. All those times normal or hard tasks did work flawlessly I just accepted it without thinking that I was lucky.

When playing Blood Bowl on either the PSP or PC I always felt that the AI was cheating since I was constantly getting all the bad rolls in the worst situations. I had a hero runner who needed to sprint 1 field to get a touchdown that would win the game. He needed a 2+ on a 6 sided die to not fall.

Ll incredibly true honestly you should only buy keys off key traders because they usually sell them at 1.50$ (only really worth it if your a starting trader or a avid unboxer) also just bind buying all unusuals for 5$ you would be suprised how many noobs fidn 1 by blind luck and will do that trade

Online gambling is illegal... I hope you didn't just get steam shutdown Jo x_x

I don't think there's enough statistical controversy on this comic yet. Shall we discuss the Monty Hall problem?

I just delete all crates as they come in. It sends an even more clear message to Valve than letting them stack up in my inventory.

[url=#user_comment_137266]@DK[/url]: VG cats did a good spin on that scene: http://www.vgcats.com/comics/images/120318.jpg People need to be more mindfull on how they spend their money.(vote with your cash not your whining) Complaining afterwards just make you look like a chump. (witch is true btw).

Ironically, statistics were invented by humans, and are (theoretically) impossible to fully understand. This is because we invented them, but our minds aren't made to use statistics and probabilities. Oh, and crates ARE gambling.

Sees people actually spending insane time on trading. SCRAP BANKING?! Quite possibly the worst of it. "Give me 2 items and I'll give you a scrap, then I'll resell those items at a higher price." Stop wasting so much time on trading and just get play the game. Pre-crate days were awesome.

I spend $5 a month on two keys. Which I save and trade for what I want.

The eternal dilemma of every Game Support Agent, ever.

"Hello, I've thrown my money at you to open 9,000 <insert random-item-giving-thing here>. I got nothing but shit. Give me a premium item NOW!!!!"

been broken by the rolls of these other players.

even if the rng algorithm used to determine scout's crate was known, the fact is EVERY player who opens a crate on that server is potentially getting the sequence-breaking 'good' result scout is looking for. His belief that the streak cannot continue forever is an illusion because it's already

But if the result is a fixed thing (one Magic the Gathering booster pack in X number of booster packs will have a super-rare), go ahead and buy the entire box. You'll get lucky on one of them.

Basically, it depends on how the item is generated.

If it is generated separately from the rest (aka: coin flips), then each generated result will be the exact same probability as the last.

... have the golden ticket, you're drawing from a reduced basket (less chocolate to mess up your chance at the ticket), and thus have a much higher chance of getting said ticket.

[url=#user_comment_137215]@anon[/url]: you got it correct and wrong at the same time For random drops, where each item's probability is generated right at creation, the odds of getting an item is the same for each For the "basket-choco-ticket" example however, you are removing tickets. Thus, if the people before you didn't

Those clusters're just individual blips in the sea of numbers.

So, the human brain tends to expect that results such as five misses in a row won't come up unless the mechanism's somehow broken. It expects hits and misses to be more evenly spreads out. But that's not how probability works. Clusters of hits, clusters of misses, and more mixed clusters are normal.

Thus, if you play games like FFT long enough, such streaks will happen, and they'll seem unusual even when all of the thousands of attack rolls we make add up to the same odds as that listed for the attack (65%-80%). Basically, the human brain mistakes probability for something more predictable.

Example: five heads in a row will seem unusual to us, but it has the same odds as heads/tails/heads/tails/tails. Exactly the same odds to get either sequence. Of course, if one considers the odds versus all other possible outcomes, it's very low.

Overall, our brains aren't very well geared for understanding raw cold probabilities. They're optimized for pattern-finding in a way that's more useful for other tasks than for statistics. Thus, there're lots of cognitive traps when it comes to statistics.

[url=#user_comment_137236]@Freelance[/url] Streaks like that happens in any game that uses hit/miss chances like that. Consider confirmation bias, and that we pay more attention to occurrences that seem unusual (instead of just looking at all of the results so far).

My whole gaming life is a lie!

@NerfNow: My sentiments exactly.

@Kaja Rainbow: Try playing Final Fantasy Tactics and given anything from a 65-80% chance of success in your blows, and missing _at least_ five times in a row.

[url=#user_comment_137203]@Math[/url]: Look at it like a roulette table and you're putting your money on a single number. Yes, it's possib;e that your number will be landed on eventually, but it's still a 1/37 chance, just like all of the rolls before it. Numbers are not being removed; the wheel remains absolutely the same.

The next coin flip is the same as the previous. The odds of 100 of the same result is vastly low, yes, but that's combining the probability of a hundred results together. It doesn't affect the probability of a single result (i.e. the next coin flip).

@Tera Online:
and that scout is counting chance of single boxex on with 'I will get smth nice' vs 'I will get crap' which looks like 50/50 (or whatever he is thinking)
I dont play TF2 but I guess probability to get some epic stuff from a single box is smth like 1-0.1% ? :D

@Nya-chan: yeah, pretty much true.
In Scout case he is counting chance of single boxes not pool of them.
AND THATS WHERE HE FAILS AT GAMBLING. :)))

tl;dr - Once you open a crate you can't consider the result as relevant for the next one, unless you considered it BOTH relevant before you opened the first one.

@Tera Online: No, the chance is lower if you consider the target group BEFORE you start.

So "I bet the next 20 crates will have at least one nice item" is valid and you have solid chances that's true, but "The last 20 were crap, so this one will surely won't be" isn't.

Learn some Math.

Interestingly enough, there are folks out there who are amazing at metagaming with spreadsheets and what have you, yet STILL get sucked into things like this.

Every toss have a 50% chance to get 'bad side' or 'good side'

But chance that you will get 'bad side' ONE HUNDRED times IN A ROW is not 50% anymore.
Chance is LOWER every time you toss the dice.

Toss the coin.
You have 50% chance to get one of two outcomes.
Each time you toss that coin chance is the same.
Your 10th attempt still have same 50%.

BUT NOW...
stop counting single chances that are the same.
Chance that EVERY of 100 try will get same result is much lower.

[url=#user_comment_137203]@Math[/url] Can't tell if trolling...

[url=#user_comment_137203]@Math[/url]: no, the odds the 100th crate will be bad is still the same. Think of it like dice. If you roll the dice 99 times and get 4 each time, the chance the 100th roll will be a 4 is still 1/6.

Classic "gambler's fallacy."
Another example is when there's a basket full of regular chocolate and one with a golden ticket.
Everyone in the room takes turns taking one out of the basket.

Probability-wise, it doesn't matter when you get your turn.

Ah, the gambler's fallacy. How I do love thee so. You and the rest of that whole complex involved in what is known as "randomly dropping"

the odds that you get a bad one after 99 bad ones is smaller than at the first crate Sir. Ofc it its gambling however if you do math you will know im right

English doesn't work that way either.