Shared publicly  - 
(Disclaimer: I'm not an educated mathematician. Feel free to double check my results and tell me if I'm wrong)

So, to explain the lottery question and the very unintuitive answer..

Every time you don't win, someone else wins. The number of people who have won the lottery once will very quickly grow significantly larger than the number of people you are. Since they are more than you, the odds of one of them winning are higher than the odds of you winning, which means that there will eventually be at least one person who has won the lottery twice.
That person has the same chance of winning as you do, so each drawing, you both have a one-in-a-hundred-million chance of winning. If neither do, either someone who has never won will win, or someone who has won once will win, and there will suddenly be twice as many people who have won twice than you are.
Because of the high number of participants, this gets iterated maaaany times before you're likely to win (50 million times on average), and the pool of winners who aren't you has plenty of time to grow.

Even in a lottery of just 1000 people, your odds of winning before someone else wins three times is only about 15%.

The odds of you winning a fair 100 million person lottery before someone else wins TEN times in total is about 50%.

This, my friends, is the power of the pigeon hole principle.
Arty Wang's profile photoMatthew Meeks's profile photoWeston Harder's profile photoLee Mei Ying's profile photo
First, hello notch, love minecraft <3
On topic: mindfucked o_o
That, my friends, is the reason I hated Statistics 1.
First? Seriously? I'd hoped we as a group would move beyond that. Guess not :/
This hurts me head. AP Statistics did nothing for me, so I'll basically ignore the question and focus on one word: Lottery.

Are the NPC villages going to have a lottery system? If you win, do you get Diamonds?
Even without being a frickin Einstein I always knew that playing lottery is stupid
ok this makes sense, yet it is hard to understand. I suck at maths - confirmed once again
That's why your time is better spent making a lot of close friends who play the lottery and will share the winnings.
That sounds about right. I remember overhearing a conversation where someone was criticised for picking their numbers consecutively (1, 2, 4, etc), because "those numbers would never come up". lol
I love the birthday paradox too. It's so counter-intuitive that probability of two people having the same birthday date is more than 50% with just 23 people, and 99% with 57 people…

Serves us right not to trust intuition when it comes to probabilities. :)
What I am interested in is-- what is the behavior as N approaches infinity? Does your probability tend to zero? Does it have some non-zero lower bound?
Yet, it is still a lot better odds than NOT playing :o)
A couple that won over 5m Euro in the Irish Lotto a few years back recently won 4.5m Euro..
I love math but hate statistics. I need to work on some imaginary numbers to get my brain to work correctly again now...
Aaron, maybe I misunderstand what you're saying, but the lottery is quite legal in the US.
Still brilliant. Still having Martinis. Cheers :-D
+Markus Persson This is why many have called lotteries a taxation of the ignorant. You are more likely to find an apple in Minecraft than win something with 1/1,000,000 odds.
It certainly is gambling, since it is a game of chance with a monetary reward. I know a lot of people that buy a loto ticket each week, but I don't think I've ever heard anyone blow a fortune on tickets.
This other condition makes it even worse.
They = Everyone who has won the lottery, not just one person who has won. You = only you (1 person).
That's not entirely true, because not everybody guesses the exact number of the lottery. The chance of winning is 120 million to one.

So if each human being on the planet played the lottery on the same week, the chances are still horrible for one person to win. 7,000,000,000/120,000,000 is 58.-33-. So with simple math, 58 people should win the lottery, but the chances of that happening is basically zero. I'm horrible at math, so please, anybody else, add something better for me...
Any chance you could display your numbers. Me and my friends (truthfully) love doing math so this would be fun to look over. Give us a break from the normal argument of how to divide by zero.
I think Notch is talking about a EU or Swedish based Lottery that is using even worse methods than the U.S. but honestly learn to play Black Jack, Texas Hold em or Roulette, I hear Baccarat has good odds as well but no exp there. I love games.
Can somebody demonstrate the 0.5% mathematically? I need formulas and numbers. Give me my fix!! :)
Easy to figure out, you just have to realise you are 1 person, whilst everyone else is...well, everyone else. It's like saying if you pick one of ten people at random (of which you are one), there's a 9/10 chance that you won't pick's just worded differently.
I have one more non-intuitive example for you. Let's say there is a test that shows if person have some disease. Results of this test are correct in 99% cases. That means there is 1% chance that test will tell you that sick person is fine and vice versa.
Now, let's take group of 1000 people. We know that 0.5% of them are sick, but we have to make test to determine which one. What is overall effectivity of this test?
The answer is: 33 %

Why only 33%?

Because 995 people are fine. Test will show 1% of them as sick and this is approximately 10 people.
On the other side 1% of 5 sick persons is approx. 0, so the test show all 5 as sick, and thats ok.
So test results give us 15 sick people. 5 for real and 10 by test mistake.
@Aaron Ahmed: Gambling is restricted in the United States, but many forms of it is legal; it's not a blanket law. Specific rules vary from state to state, but lotteries are legal in all states.
+Jakub Sobel Your math is off 5 percent of 1000 is 50 so 950 not sick. I am interested in the process as I have a condition which 5 percent of the population has hyperthrombosis. My wife +Cassandra Brown is the math person and she said by your standards the test has an 80 percent success.
I've been enjoying this problem with my roommate. We're trying to generalize it with T's and W's and using that stuff we learned back in probability and statistics. I even cooked up a quick computer model. I'll let you know if I get a really pretty maths result.
Did you get this from You Are Not So Smart? Its a great blog that talks about stuff like this. (to clarify: I'm not affiliated with the blog, I just think its cool.)
Does this mean we get to gamble in minecraft now?
Too bad probability is purely theoretical (you will seldom see it work in real life)
Chase S
Nice, you didn't mention the chance you have of winning 3 times before someone who is not you wins once.
Here is the way I like to think about this problem: Imagine it was just you and one other person, and the question was whether he would get 3 wins or you would get 1 first. Any lottery where someone else entirely is the winner doesn't matter, so you can just consider the ones where one of you won. You both have an equal probability of winning, so you each have a 50% chance of winning first. Then you each have 50% chance of winning second, then 50% third. Your opponent gets 3 wins before you get any if he wins all three times, which happens with probability 1/8. So any one person has a 1/8th chance of getting 3 wins before you get any, and there are 999,999 other people, so clearly it's very likely that someone else will beat you 3 times before you win.

The actual number is not quite as simple as 7/8 ^ 999,999 because if you win against one person, you are fairly likely to beat someone else as well, but this reasoning gives good insight into why you probably aren't going to win before someone else does three times.
+Chase ... That one is easy, you have to win three times in a row at the beginning, which happens once out of every 1,000,000,000,000,000,000 times.
Ok, my bad, it should be 0.5% sick people. Then there is 5 sick and 995 fine people ;) Corrected.
Well, assuming you never die, yes. But if you played every day for 50 years, the odds are only 1 in 5480 [ (1-(1-(1/100000000))^18250)^-1 ] that you would win it even once.
the chances of making an indie game that the internet goes nuts for seems like an easier way to get a million dollars, I think.
I take a class in discrete math and Notch starts talking about it. Sometimes the world just seems to like me.
1. That isn't the pigeon hole principal.
2. The probability of you winning is the same every time so nothing else really matters...
Well, yes, his solution only works for immortal spherical chickens in a vaccuum. :P
This is why computer science/applied mathematics is awesome
+Ignacio Parentella You can see this situation in real life. There have been several people who have won the lottery multiple times. Just Google: won lottery multiple times

Also, I wrote a C++ program to simulate millions of drawings between the posting of the problem and the solution. I was bored and could use the practice. With the pseudo-random number generator, winning the initial bet appears to be fluctuating around 4.7% out of 1000 loops to 3 wins (typically takes 500-3000 drawings per loop, I should have it keep track of the min and max in each test).

EDIT 3: Implemented a Mersenne Twister for the random number generation. Probability has dropped to as low as I initially expected. I'm somewhere in the middle of a 1000 loop (each one taking an average of around 300k drawings) test for 100 million participants to 3 wins... and I have only noticed one SUCCESS scroll through the console so far.
Sometimes, in a lottery, nobody wins.
All well and good indeed, but given those odds of winning, where exactly have I left my other shoe?
Actually, the pigeonhole principle hasn't much to do with your problem, really.

It's basic probability, but I hate that stuff so much that I can't tell for sure how to work this stuff out.
"That person has the same chance of winning as you do, so each drawing, you both have a one-in-a-hundred-million chance of winning. If neither do, either someone who has never won will win, or someone who has won once will win, and there will suddenly be twice as many people who have won twice than you are."

This doesn't make sense.

The rest of it does.
Notch this is why you're a part of Mensa.
I don't see that as unintuitive at all. All that this is saying is that that even if someone else wins, you're no more likely to win than before. No probability changes here.
But you could be lucky and be the guy who wins 3 times.
so there will be a lottery in Minecraft?
+Ben Wiley - That makes sense when you think about it like this.
You are 1. You = 1.
Person who has won = 1.
Person wins a second time = 1 + 1 = 2.
Person wins a second time = You * 2, or 1 * 2.
There are now twice as many people who have won twice than you are.

Written math doesn't follow regular grammatical rules, so it is intuitively harder to follow if you don't parse it out mathematically. Parsing it out grammatically will only end in confusion, and at times, incorrect calculations.

Another good grammar vs. mathematics example is:
Seven and two is nine.
That is incorrect grammar, because you would say, "Seven and two are nine". Just like saying "apples and raspberries are red". Apples are red, and raspberries are red. But seven is not nine, and two is not nine, so the grammatically correct version is mathematically incorrect.

The moral of the story: when reading mathematical explanations, don't use your understanding of grammar to understand it; instead, use your understanding of mathematics.
I think I need a Kahn Academy video to explain this...
either way the lottery (in belgium at least) sponsors a bunch of organisations (including a lot which work for charity), so playing with the lottery is giving to charity with a (albeit small) chance of winning stuff
'blah blah blah I like pigeons' that's what I read.
Eric L
intresting. :) thank you notch for this little computer science lesson.
This depends on how the lottery is set up, there are various kinds of lottery, are we talking the kind governments use for rounds of conscription, where once one has one they are taken out of the pool? Or ones where people choose x numbers and have to match them? Then there are ones where every entrance is a unique number and X random numbers are chosen each month as winners.
+Michael Foley Lottery just means drawing lots for prizes at random (well, the prize could be you are the first conscripted, like in the US conscription system).
I'm not a pro but I feel like something's wrong here.
Just a little test I made to picture it : 6 people, 1 dice. Everyone is associated with a number. We roll the dice three times and see what happens. I'm number 1.

Possibilities : 6x6x6 = 216

Cases where I win at least once : 91.
= 42% chances of "wining something"

Cases where I win all 3 times : 1

Cases where 2,3,4,5 or 6 wins 3 times : 5
= 2.31% chances of not wining before someone wins 3 times

Cases where 2,3,4,5 or 5 wins 2 times (in a row or not) : 65
= 30,09% chances of not wining before someone wins 2 times

If I'm equal to every other players... well, there's no reason why I'd have less chances to win once than I'd have to win several times.
If I read what you wrote correctly, you assume that somebody wins every time numbers are drawn. This assertion is false.
Yeah, sounds about right to me. But I'm not that much into statistics and stochastics, though. Real (!) math that engineers deal with is more what I'm good in.
I was skeptical, so I wrote small program to simulate this situation. I made a lottery with 1000 "people" that ran until a) person number 1 won, or b) someone won 3 times. I then ran this lottery 10,000 times, to get a good average. After about a minute, the program spit out these results:

Times you won before anyone won three times: 1494
Times someone won three times before you won once: 8506

Hey, what do you know. That's within tolerance of the given number of 15%. Probability is weird.
Shawn Haggard: his assertion that someone wins every time is not false. Its the very problem statement: "Let's say we have a fair lottery where a hundred million people get one ticket each, and a winner is drawn at random each day."
Following ruby script runs the Monte-Carlo approach:
number_of_participants = 1000
number_of_runs = 1000000

you_won, they_won = 0, 0
number_of_runs.times do |i|
you_win, they_win = false, false
won =, 0)
until (you_win or they_win)
winner = rand(number_of_participants)
won[winner] += 1
you_win = true if won[0] > 0
they_win = true if won[winner] == 3
you_won += 1 if you_win
they_won += 1 if they_win

puts "You won #{you_won} out of #{number_of_runs} (= #{"%6.3f" % (100.0 * you_won / number_of_runs)} %)"

You won 154904 out of 1000000 (= 15.490 %)
(sorry, but G+ keeps removing the indentation)
That's why I'm unable to play fortune games
That first part is not true. Every time you don't win it does NOT mean that someone else wins.
uuuuuuuum........................................ OK
For example, you can buy two tickets for the same jackpot, lose both, but there's still only one winner of that jackpot.
Guys give him brake its sunday time to have a weekend to yourself i am sure he will be working on it 2morrow
I ran above script with 10,000 participants and one million samples and get: "You won 74003 out of 1000000 (= 7.400 %)"

With 100,000 participants: "You won 3478 out of 100000 (= 3.478 %)"

With 1,000,000 participants: "You won 1610 out of 100000 (= 1.610 %)"
My simulation (after fixing the random number generator) of the problem as described (with the full 100,000,000 participants) just finished and produced the following results:

Loops = 1,000
Failure = 997
Success = 3
Percent Successful = 0.3%

The rounds ranged from 33,731 to 748,652 drawings until someone got 3 wins... with a total of 346,400,666 drawings simulated over the 1,000 rounds in the test.
thanks notch, for mind fucking me hard.
A little blog post of notch, millions of mind blowns.
Most people, once they win the lottery, don't buy more tickets
Makes sense to me. Then again, consider this way: If you won the lottery once, the number of people who haven't won the lottery yet is far greater than the number of people who have won it once, and the number of people who have won it once that aren't you is far greater than you. Personal chance is always lower than collective chance.
I hate getting involved in these little bitch sessions - for or against - but, I would just like to point out that this is, fundamentally, a question about probability, not gambling.

Has it occurred to those of you who feel Notch would be serving you better by working on Minecraft, that this may, in fact, be related? He may be working out a probability function for Minecraft, or he may have run into this interesting little titbit whilst working on probability functions for Minecraft.

Just saying that, maybe, you should consider things a little more, before you take an aggressively negative standpoint, and shoot your mouth off. (well, technically, "shoot your fingers off", but I can't see it catching on as a turn of phrase)
I had someone once tell me that they were not going to buy a ticket for a 'superdraw' because more people would buy a ticket for those, and so their odds of winning were less.

I boggled at them, and tried to explain that it didnt matter if one person or a million bought tickets, it didn't change the likelyhood of winning.
The only thing that would change is the dividends, as the chances of multiple winners would increase with more tickets sold.

The thing that really upset me was that this person was a partner in one of the country's top law firms, and I a simple accounts clerk.
Joe L
There's a lot of misconceptions about probability - like people who think they have a better chance guessing the same number every time
Pigeonhole principle is cool, but proving it correctly without induction can be a nightmare.
You just blew my mind, good sir.... But I love MC :D
Sweet... That would mean I will win the lottery 3 times before you do!

HAHA I'm gonna be richer than Notch!
Good thing for me that I don't believe in mathematical statistics, excuse me while I go get rich off of lottery tickets XD
Discounting the fact that most multimillionaires don't invest money into lotto tickets of course, thus after winning or already being part of that portion of society, that they wouldn't compete - thus you are still in the same pool of people who have never won a major winning.
This is the pigeon hole principle in part.
The number of times people have won against the number of people who have won.
Let Mac OSX Lion get itself sorted out first.
nice makes perfect since and its pretty cool
its actually higher. Someone who won the lottery and will play again is likely to buy more than one ticket and likely to have the money to buy more tickets than you.
Somehow gambling principles, no matter how true they are... appear to have the aspects of half-truths, or bent reality. There's always an angle left out,,, like whether or not a there might be a series of lottery's and how that might affect the probability. There is also the problem of competing probabilities when considered from different sides.
Taking Discrete Structures II in a few weeks when school starts back up...

Thanks notch for ending my summer early... lol
So uhh... does this mean NPC villages will have a lottery?
Hey Notch, can you get back to me on this in about six months? I'm not taking statistics until this coming semester. Kthxbye!
Lottery only makes sense if the expected payoff is greater then the cost of a ticket. Then it makes sense to buy all the possible combinations to ensure you hold the winning ticket. This situation does occasionally happen with roll over lotteries and there are gambling operations that have successfully carried out the procedure.

Or one can figure out how to game scratch off tickets to always win, which is sometimes possible.
Except that you don't take into account the psycho-social ... The winner of "huge price" lottery replay a little or no at all !
Thank you "satiety" and regulation by social pressure !
Which, by the way, is idiotic for the winners, because with the money, it's easy to industrialize participation in the lottery !

Another sleight of passes for those who are still fried on following your presentation: To find someone who has won 10 times in a row to coin flipping, you just have to hold a tournament with 1024 participants initially ...
A twist on this one: What are the chances that someone else in the world, anyone else, will get struck by lightning three times, before you get struck even once! Now it doesn't sound so far fetched, does it :)
Probability. How beautiful and in this case COMPLETELY unrelated to minecraft. GET BACK ON THE ADVENTURE UPDATE!!!!!!!!
+Troy Ablan. Well hopefully after getting hit once they might learn from it and not go dancing in a thunderstorm... but then there are those who don't and continue... now your adding human quirks to the equation xD makes it slightly harder...
Sounds about right, the only thing that gets me still is the Birthday Problem. I understand how it works, but it still never ceases to amaze me.
James B
However it does ignore that some people who have won, won't play in future rounds. The odds still aren't in any individual's favour though even if that's taken into account.
Ok. Understood ! But I don't play lottery! I'm a geek so I play Minecraft
"Chance of winning = X / tickets sold" where tickets = 100 million

Chance of winning in a fair 100 million person lottery with 0 tickets purchased = 0.
Chance of winning same lottery with 1 ticket purchased = 1/100,000,000

1/1,000,000 is an infinitely larger chance of winning than 0. Of course, infinitely larger than nothing is an interesting concept as 0*infinity is undefined.
The only way to know your chances of winning the lottery is to calculate the number of combinations. For exemple with Euromillions: ((50x49x48x47x46)/(1x2x3x4x5))x((11*10)/(1*2)) = 116 531 800 combinations. Thus, you have 0.000 000 9% chance of getting the full combination per ticket.
@those telling Notch to get back to work, where do you think they get the witty comments thats sit up near the Minecraft logo in the main menu?
It depends on which lottery you're playing. The mat is different for each one. 
There is always 50 % chance for everything: You win, you lose.
You die, you live. You make good video game, You don't make Minecraft
Ok, now I imagine Minecraft villagers playing the lottery... no... just, no.
If you do not want someone to bring a bomb on a plane, bring one yourself. It is very unlikely that someone will decide to bring a bomb on your plane out of the millions of flights in the world, but it is even more likely that two separate people without any connections whatsoever will decide to both bring bombs on the same plane. Every situation has some way of creating false probabilities that seem to workout. That is why statisticians are considered to be the best liars.

But then there is also quantum mechanics in which the probabilities involved are mind blowing. But that is on a whole other train of thought. Just try to stay away from cats and boxes and you'll be fine.
If statistics weren't so mind numbingly confusing, the lottery probably wouldn't be so popular. My brain hurts...I might go buy a ticket to ease the pain...
+Joni Finne , that statement is fundemantally flawed. There is two outcomes, but it doesn't mean that the probably of both outcomes is 50/50. This would imply that they would both occur with the same frequency. (Which it doesn't)
This is exactly why I don't play the lottery. I know that the chances of winning aren't good and it's not even worth trying. Besides, some people, like the ones that won before, buy tens to maybe hundreds of tickets each time to increase their chances...
Every time I don't play, I win the minimum cost of one ticket by virtue of having not played. I am an extremely successful not player. 
I am not an expert in statistics myself, but it seems to me that you are mixing up probabilities on the level of the population (100 million people) with probabilities on the level of individuals. In statistics you cannot simply draw conclusions about individuals based on findings on the level of populations. Another example: if there is a road where 1 in 10 people that pass that road crashes, it does not mean that you personally have a 1 in 10 chance to crash on that road.
I imagine a group of creepers gathered in a circle.. A conversation ensues: "SssSSSssSSSss?" "SssSSSsSSsssSSS" "SSsSSSSssss!" One throws a dice, number five comes up. "SsSSSssSsSS" one says and points towards a glow some distance away where Steve hides in a wooden house. "SSsSSSSsSS" number five replies and shuffles off. Five minutes later a lout "OH SHI..." and a <BOOOM> is heard. The rest of the creepers jump with joy and sing "SSSsSSSssSSSSS SSsSSSsssss".

Happy days.
Stop writing pointless stuff into google+ minecraft isn't working! I forced the update 10 times now but it won't work! I want to know what I paid you 15€ for. Windows 7 Home Premium 64 bit - FIX IT.
That's assuming you are allowed to win more than once. The question said wine ONCE before someone wins three times.
See this is why i'm a modeller/animator, math is ridiculously hard to me i get in a sweat when asked the simplest multiplications sometimes, kinda makes me glad i didnt do computer science haha.
+Lukas Hempel, You paid him 15€ to help beta test his game in return for a free license of the game proper when it is released. If it truly is an issue, he will "FIX IT," and not because you shout-typed it in a comment on a G+ post. In the meantime, I'm sure you can find something to do.
+Lukas Hempel First, manage your obviously over the top anger. Second, describe your problem, "FIX IT" doesn't help when no-one knows anything about the problem in order to help diagnose and propose solutions to said problem. So, what's not working with minecraft? When and where does it happen? What do you do to get the problem to occur, etc...
Even on Google+, people are idiots. You expect the man to ONLY work on minecraft? Do you work 168 hours a week?
at least on here, you usually get peoples names, now the whole world knows lukas hempel is a fool
Of course. Like how I know 9/11 was a hoax and that its all Obama's fault.

(I expect the man to have a life, a 40 hour work-week is good enough for me.)
I had this problem with an employee lottery thing we did years ago, changed the algorithm to allocate points based on how long you were hired for and how recently you had won. This way everyone would basically win once before you had a chance to win again. Employees did not understand this whole concept and some were complaining the thing was "rigged". The funny thing is, we had to rig it so people didn't think it was rigged.
Only 10% Michael? I envy your focus.
Educated Mathematician, how far from it am I? A lot!
Minecraft works bad on Mac OSX Lion. ;'( Please Notch fix it. Blocks placing goes really slow in SP.
i cant comment on the other post cause its full, but i got this error when trying the launcher loader 2.0 on WinVista SP2 - "This XML file does not appear to have any style information associated with it. The document tree is shown below."
Here's the big question: did you buy a ticket? Your logic seems sound, however to quote Han Solo, "Never tell me the odds." The way I see it, none of your equation is valid unless you buy a ticket, and at that point there are only 2 possible outcomes; you win or you don't :)
Two possible outcomes, but not equal chance of each!
@Chris He's probably not doing this during his work time! Can't a guy have some rest? Lol
@Chris, we didn't pay for him to finish the game. We paid for the game as it was at the time we paid. Didn't you read? :)
Joe Le
Who cares about math when you win the lottery :P
Assuming 1 person always wins on each draw.
Consider that the probability of not winning is 99,999,999/100,000,000.
So the probability of not winning n times is (99,999,999/100,000,000)^n.
The number of draws until the chance of you not winning has been reduced to 5% is therefore:
log(0.05)/log(99,999,999/100,000,000) = n.
n ~= 300,000,000.
So in order to be certain that you will win (95% confidence) the number of winners is huge, 300M.
Even if you make the crazy assumption that everyone else wins the same number of times each, everyone else will have won 3 times before you can be certain of winning yourself.
You have to compute the expected value of the feature, and the variance or distribution. The lottery win is uniformly random I suppose, but this statement "Every time you don't win, someone else wins. " is not the same way the US Lotteries work. Ball drawing number schemes often do not have a winner, and when "123456" wins, the prize is split among many people.

"Since they are more than you, the odds of one of them winning are higher than the odds of you winning" This is also not true, until we begin to approach the number of players, or half the number of total players. That would require millions of wins. Sometimes a quantification is a good thing in a deductive argument. If this premise needs millions of wins to come about, it wont be happening any time this century (5200 weeks).
+Chris Christner I think you are over reacting, as he is clearly doing work. People who expect him to do stuff 24/7 are just ridiculous. It makes no sense, do you work 24/7 at your job?
@Aaron Ahmed, you derp. The "first" comment was using it in the order of statement sense. First, hello. Secondly, goodbye.
But if my lotto ticket is box and I don't know the numbers on it. Then what? I'm a both a winner and a loser?
Hey you Chris guy, here's a few things you're doing wrong:
Knowingly overreact.
Judge bug fixes as insignificant
Think that buying an unfinished product gives you the right to demand it be finished on your schedule
Think that anyone lets notch fart
Criticize and celebrate someone with the same breath
Assumed that notch doesn't want to finish the game

Seriously I'm getting pretty sick of all you children whining at notch to finish his game. You look like fools. Grow up and go about your business. 
Very interesting!

You could also just put it like this: If you buyone ticket, youll have that one chance (whatever that percentage may be) and if you buy two tickets you have exaclt doubled that chance, if you buy 3, 4, 5.. and so on, so the percentage must be much lower than 15% no?

hahahah anyways, good braining!
Hey Chris, we got pistons - but I do not get your point ...
Ya pistons where nice, but from what I understand he took something from a mod someone else did. And other than that in the last few months, power rail tracks...yay. The adventure update sounds like it will be worth the update, I guess. Clearly minecraft has way too many blindly loyal fans for me to say anything bad about it. I regret saying anything.
+Chris Christner I am not saying that I don't want more updates, but I am saying that the rate at which we get updates seems fine to me. It is in beta, and is full of bugs. I dunno if you have ever been in any other games beta, but thats pretty much all the updates are...bug fixes. I understand you want more content, you feel you paid for more stuff, and I am not saying "No you are wrong, Notch has a huge dick and I love it" I am just saying that there are updates happening, and the next one looks huge. I just don't understand why everyone qqs about this.
WOW +Tristan Bowles thank you, this is one of those times I should have just said nothing. This is my first time buying into beta, I usually wait till one or two patches comes out for a finished game before I go for it. Really wasn't trying to give Notch shit, just wondered why he was posting about the lottery. Thanks for having a different opinion but not being an ass.
Holy crap. Did someone just work out their differences on the internet? I didn't know that was actually possible...!
This is my favourite kind of maths, The kind you're not sure you've fully understood, but have still had your mind blown...
+Jordan Becks Not sure if they actually worked it out or if Chris just resigned because he was tired of fighting "fanboys".. I'm agreeing with Tristan though, Notch is doing some great work but I can see how some people think he's been a bit on the back burner lately. I guess we all just got used to when he pushed out new features by the hour (with the bug-rush to go with it) :P
Minecraft is awesome. Statistics is not.
+Chris Christner Things you're still doing wrong:
Reverting to vulgarity and insults because you have no legs to stand on.
Continuing to assume that bug fixes are insignificant to the big picture (it's called professionalism)
Assuming that your earlier posts had anything constructive in them.
Again with the vulgarities.

Move along plz.
You have enough money. If you win on the lottery.. share the amount with the entire minecraft community. That way we all win :p
It's got to be more than >1/2 %. With a hundred million people the odds of someone winning the lottery 3 times is 1/(100,000,000)^3 .
Unless you buy scratch-off instant win tickets like I do. Then all this goes out the window.
Wow... that's really messed up! Did you come up with this or share it from somewhere? I love this explaination, can I share it!
The trouble with this scenario is that with modest expectations attached to it (e.g. - 50M is the only prize, 10M participants, $10 tickets, weekly lottery), it'll take over a hundred thousand years to get to the point at which one would expect the population of winners to substantially exceed the number of losers.

And "by the power of compound interest", putting away $520/year, for any interest rate higher than thousandths of a percent, tends to make calculators blow up (e.g. - blowing past 10^38, which is commonly the limit of floats). Thus, saving your $520/year would make you "infinitely" more money over that period of time than waiting to win the lottery.
It makes sense and isn't very surprising. An alternative example from Wikipedia:

We can demonstrate there must be at least two people in London with the same number of hairs on their heads. A typical head has around 150,000 hairs, therefore meaning that it is reasonable to assume that no one has more than 1,000,000 hairs on their head (m = 1 million holes). Since there are more than 1,000,000 people in London (n is bigger than 1 million items), if we assign a pigeonhole for each number of hairs on a person's head, and assign people to the pigeonhole with their number of hairs on it, there must be at least two people with the same number of hairs on their heads (or, n > m). In the worst case (the case with the fewest overlaps), there will be at least one person assigned to every pigeonhole; even then there are more people than pigeonholes.

This pigeonhole principle grows as the population of events (lottery drawings or people in the Hair-Counting case) grows.
Curse you, you're not supposed to think about maths in summer xD
But you also have to remember that people are constantly joining and leaving the lottery over time.
"dice have no memory" or something like that... Every time the lottery is held it's a brand new day with no knowledge of the presiding days... It is reset and it starts again.

That said how unlikely has something got to be before statistically it's simpler to describe it as impossible? In the UK the chances of you winning the Lotto is over 17 million to one! that's damn unlikely, ~So unlikely as to be impossible, and that is why I don't do the Lotto!!!
Alex, Markus it taking as a start point the rules in the post before this one: "a hundred million people get one ticket each" + "The winner stays in the pool even after winning" means there's no one leaving (actually, you are right in a certain sense: it doesn't explicity says people not wining stays in the pool, but for the sake of simplicity...)
hmmm I think i understand or do I wait I need to re-read this
plausible and sounds right... isn't that hard to understand, is it? >.>
Okay... so imagine you're Notch. If you post 1 crazy random question, where does this get you in Minecraft development? 
the moment somebody wins 9x in a row imma getting a lottery ticket :D
it's a lottery inside a lottery inside a lottery!
Hey Notch! I'm Ashe. It's a pleasure. =P
Congratulations on Minecraft's success. Keep up the great work. =D

About the topic: I suck at math and math hates me. I don't know very much about odds, but I still play on the lottery. It's the dream/hope, man... We never know.
"Winning" is based on your own personal definition. If "Winning" to you is getting the jackpot of however many million of your currency, than your chances are quite close to nothing. One can wish. Or, if winning is getting say a $5 check from getting two of the numbers right, then yes, you have quite a high chance of winning. Then, there is luck. Someone could win the lottery ten times before you win it once, but unless they're hacking, than you have exactly the same odds stacked against the both of you.
So, if I lose, someone else wins, and the more someone else wins, the more likely I am to win. Does it follow that the more I lose, the more likely I am to win?
Though on the other hand, you're 85% likely to win three times before someone else wins once. Right?
For anyone interested in the actual math behind this problem, this is what I've got so far:

I've solved this problem for someone else winning twice instead of three times. The answer is 0.01253280805026440%, or about one percent of one percent.

Now for the math:

[k=0 ∑ n] (expression) = sum from k = 0 to k = n of expression
[k=0 ∏ n] (expression) = product from k = 0 to k = n of expression

let p = 100 million (the number of people)
let e = 1/p
let f(n) = e * [i=1 ∏ n-2] (1 - i * e)
solution = [n=1 ∑ p] ( f(n) )

For anyone interested in an explanation, either post here or send me a message. The next step is to extend this problem to winning three times.
These statistics are only relative. From my perspective this may be true, but if I play AND my friend plays, we could look at it from my friend's perspective. The odds of my friend winning (1,000 person lottery) before I win 3 times would be 15% and the odds that my friend wins a 100 million person lottery before I win 10 times is 50%. Provided you typed 'someone else' rather than 'someone else in the group' on purpose, these statistics are true both ways, thus invalidating themselves.

Of course I've never taken statistics and I think I understand where you're coming from, with my knowledge at this point in time, I don't agree with them. :P Use the + or @ thing if you want me to see your reply, I'd love to learn something new.
+Zach Butler I'll give it a shot.
When he said "someone else" wins 3 times, he did in fact mean "someone else in the group." The answer makes sense because "the group" includes a lot of people. So you can't switch perspective like you did; in other words, it's wrong to say "the chance of my friend winning before I win three times is 15%". It's not "I", it's "one of the other 999 people".
Bryan Schmidt. Actually, that is quite incorrect. "Red" is an adjective. "Nine" is a noun. Therefore, your analogy about apples and raspberries being red makes no sense. Furthermore, you're not using the correct syntax for mathematics. The most basic version of what should have been said is as follows: "The sum of 2 and 7 is 9."

The moral of this story: Learn English.
recently read something about the pidgeon hole priciple, though not demonstrated by using a fair lottery...
But if I am the person who wins 10 times?
Then, you sir, are awesome. And rich. Awesome and rich. And lucky. Plenty of lucky too.
+Eoin Crafferty Grammatically, saying "seven and two is nine" is equivalent to saying "seven is nine and two is nine" which is incorrect and not what you mean. Both of your final versions of this statement are BOTH grammatically and mathematically correct. Whereas the original was only mathematically correct. "seven and two are nine" is equivalent to "seven plus two is nine" is equivalent to "the sum of seven and two is nine".

The usage difference between are and is depends on what and how it is being acted upon. While most people would assume that "seven and 2 is nine" is the grammatical equivalent of 7+2=9, this is mainly due to knowledge of the context and implied meaning. Whereas the express meaning is actually nonsensical.

To be more specific it is because of a difference in the subject of both sentences. In "seven and two is nine", "seven and two" is the subject. Whereas in "seven plus two is nine" and "the sum of seven and two is nine" only "seven" is the subject. Therefore, the two sentences have, grammatically, different meanings.

I will, however, agree that his apples and raspberries statement wasn't actually what he was trying to get at.
but if you don't add the factor of which the people who only entered once you have a chance of 72% of winning before someone else that one already reaches there tenth time of... Duh WINNING!!!!!! lol but trust me i may be a 7th grader, im in 9th grade math!
If someone just said 'pigeon hole principle' I would have thought it was something entirely different.
This is why I don't play the lottery. >.>
I completely understand Markus, but you've got some people who are rich and buy 30 tickets so its a 30% chance of winning for them and we get how many chances...... 0
the good news is i believe once someone has won the lottery, they are far less likely to continue playing.
Brian woodhouse. What about the load of people who start every week?
Mr. persson, you have broken the fourth wall.
And thus the all mighty Notch created the fourth dimension!
the fourth dimension was already created, idiot. Lisa Simpson voice
aw damit I knew i was missing out on stuffz D:
Mathew risaldi-
Oh well, still, someone has to win. And every person has an equal chance of winning. Sort of.
He said 'this my firends' w00t!, i am his friend now?
I see, so having a pidgeon poke holes in my sheet increases my change to loose?
I am dumbfounded by your logic. Yet again you cease to amaze me Markus..
i <3 notch ... so much more now XD
I got a msg from u won 5000000GBP....can anybody tell me it's true or fake...any 1 got this type of msg?
never met you before, 1.2.4 is awesome,both you and jeb is awesome, honestly.
there were pandas back then, so was monkeys.
they just did not know back then, honestly.
enderdragon vs giant, enderdragon wins. prove it. honestly.
trying one mod called more pistons, add one of them, one of them is piston with a huge air time, see for you self honestly.
add it as a non mod to make people buy of this game.
turn more pistons mod into an real minecraft piston series.
I'd always think about it by you'd have a better chance to be struck by lightning than win the lottery.
i was wondering whether you could check out the more pistons mod an think about adding it to the game it would really help redstone fanatics out with lots especially the gravitational piston the possibilities are endless!!!!!
and I was also wondering what you think of fallen kingdom by captain sparkles
markus it wont let buy minecraft );
Why does everyone think this is hard? I'm 14 and understand it perfectly.

the odds stay equal the entire time, unlucky me.
Kind of sad when you think about it. But from someone else's point of view, it's the same for them too.
markus please can you work on getting Minecraft on the ps3 if you can't no big deal but i would love that
Yes, i want herobrine back and please can you work getting Minecraft on the PS3.
But from the other people's point of view, another person might win 10 times before they do.
You should have Minecraft for 3DS, or maybe Wii.
Your a creeper,hah!Sssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssss
Notch help i log in but it says failed Help love ssbbfan1122
Now there are mutated creepers! Run for your lives!
+Daniel Nickson Dude, unless it happens at least several times in a row, don't bother the person who created minecraft. And even then, if it does happen several times in a row, ask someone else.
hello markus I'm your biggest fan and I want to ask you a favor
That is such a great game you made!! WE LOVE MINECRAFT!!! also if your reading this:
emerald tools please!! 
This is most of all correct but depends if it is any winning or jackpot
+chris hopman Technically, infinity uses boolean logic. Its either infinity (True) or not (False). You cannot approach True from False. However, if we take a very large number, the probability of someone winning twice will decrease.
Is ur Xbox360 user name xXxpaulyxXx? My Xbox name is jekloe2004
For the xbox360 Minecraft tu16 can u add the spider jokey
Вы продаете Minecraft. nooooooooooooo
Add a comment...