Lotto Numbers

  • Thread starter Thread starter ZZII
  • 89 comments
  • 3,314 views

Who has a better chance of winning? (read post)

  • Person A

    Votes: 6 11.8%
  • Person B

    Votes: 13 25.5%
  • both the same

    Votes: 21 41.2%
  • i don't care

    Votes: 11 21.6%

  • Total voters
    51
It works. Occasionally it doesn't - usually the Yahoo! section, actually - but they patch it from time to time. Thrasher made a comedy Beavis & Butthead skin for it, which I populated with THE most ridiculous sounds of all time ("SAVE CHIP!").

It's free, it unclutters your desktop if you've got a load of IM programmes. It works. Meh. :D

[/off-topic excursion]
 
Originally posted by Famine
I'm willing to stake money on it. Are you?

Let's just pause for a moment, in amongst the "Face it Famine, you're wrong" and "No-one can be right all the time", and think for a minute.

Why is it I was voted the "Most Knowledgeable Member"? It's not because of what's in my head, that's for sure. Could it be, possibly, that I check all my facts first, in books? I wonder where this puzzle came from... Hmmm...

Jpec07 - you at least have my MSNIM. I'd be more than happy to go number guessing with you. Let's make it any number between 1 and 10, to bring the odds down to a little nearer being in your favour, shall we? I'm pretty sure that you can trust me to not cheat - given my extant reputation on the GT3 board - but you don't have to if you don't want to. I can PM you a list of the chosen numbers, with a read receipt so I know you haven't read it, before we start.


Oh, and ZZII has it SPOT on.

I'll stake some money on it too. By the way, where did you hear of this riddle, Famine? It is by far the best mathematical riddle I've heard in a while. There are no good riddle web sites.

Okay anyway, next challenge. This is a relatively easy one that I'm sure at least a few people have heard, and the answer is pretty indisputable:

You have nine coins. One is a counterfeit and weighs slightly more than the others. You have an even balance. How many uses of the balance do you need to find the counterfeit coin?
 
I was going to put the answer in here, but I got 3 pages out of mine, so it'd be unfair on you :D

I'll see if I can find the original riddle...
 
Three, I think.
Five on each side.
Weigh two out of the heavier five. If one side is heavier, that is the counterfeit coin, if not, try another two. If one side is heavier on this weighing, that is the counterfeit coin. If not, the remaining coin is counterfeit. Is that right? It's late, and that's the best scenario I could come up with.
 
Originally posted by Famine
I'm willing to stake money on it. Are you?

Let's just pause for a moment, in amongst the "Face it Famine, you're wrong" and "No-one can be right all the time", and think for a minute.

Why is it I was voted the "Most Knowledgeable Member"? It's not because of what's in my head, that's for sure. Could it be, possibly, that I check all my facts first, in books? I wonder where this puzzle came from... Hmmm...

Jpec07 - you at least have my MSNIM. I'd be more than happy to go number guessing with you. Let's make it any number between 1 and 10, to bring the odds down to a little nearer being in your favour, shall we? I'm pretty sure that you can trust me to not cheat - given my extant reputation on the GT3 board - but you don't have to if you don't want to. I can PM you a list of the chosen numbers, with a read receipt so I know you haven't read it, before we start.


Oh, and ZZII has it SPOT on.



Yeh well i only joined like this week and anyway, If you got it from a book then the book is wrong because:
With your explaination if you were to choose the right door would you have a greater chance if you changed or not? If you say you should stay you're wrong as you've already proved its better to change but if you say you should change you're therefore saying they are equal (and opposite) thereby contradicting yourself.

Ok as for the coins: well one way is to weigh 2 random groups of 4, (if they're are even you've already got the counterfeit coin, the one not in those 2 groups.) if one group is heavier you get that group and divide it into pairs then you get the heavier pair and find out which one is heavier. Therefore minimum of 1 if your lucky otherwise its 3
 
Originally posted by Crayola
If you got it from a book then the book is wrong

Sure it is... Let's examine the possibilites here... Text books, governed by publishing guidelines to present only factual information, written by mathematicians, are wrong. Or you are. Hmmm.


The answer to the second riddle is "at least two" - but normally three. Timmotheus's solution was correct - although you can weigh two pairs of the 5 coins the fake must be in. If they are even, then the other coin is fake (two weighings), otherwise you must weigh the heavier pair against each other in the third and final weighing.
 
Originally posted by Famine
Sure it is... Let's examine the possibilites here... Text books, governed by publishing guidelines to present only factual information, written by mathematicians, are wrong. Or you are. Hmmm.


The answer to the second riddle is "at least two" - but normally three. Timmotheus's solution was correct - although you can weigh two pairs of the 5 coins the fake must be in. If they are even, then the other coin is fake (two weighings), otherwise you must weigh the heavier pair against each other in the third and final weighing.
Isnt there 9 coins, or is it nine coins and 1 fake?
either way its 3 at most. But if there are nine all-together and you weigh 2 groups of 4 and they're the same then the non-included coin is counterfeit. As for the other one i dont really care but tell me then if you originally chose the right door what would you do?
 
Ok........ sorry but I just read this thread and I feel that both of the ppl playing lotto have absolutely the same chance. Im going to try and explain my reasoning.

Two ppl are asked to choose one card from a deck of 52 playing cards on a certain day, mind you these cards are thoroughly shuffled. Lets say the first person always chooses the left most card and the second person always chooses a card at random. On another day they are asked to do the same thing..... choose one card from a deck of 52 thoroghly shuffled cards.

You never take away cards and as with the lotto you face the same number of numbers each time.

How could the second person possibly have a better chance than the first? Each week or whatever it is going to be 1/52 chances for both ppl.
 
To me, it seems as if the door situation is set up. I think part of it is true, and part of it is false. Say you choose the middle door, and the guard tells you its not. Hell, he even opens it gets ripped alive by dogs and is carried off screaming and yelling and leaving a nice blood streak into the middle door, reminding you not to choose that one. Unless you have a death wish, of course your going to change your mind. I find this is where changing your mind is a good idea. But now your left with two doors in mind, and each with its own possiblity.

I may of missed the entire point the door riddle. However, even if you do leave it set up how Famine has it, it just doesn't make sense to me. Why would you be considering a choice that is completely illogical? Each door has 33% right? Thats for three doors, and since each will have a 33% chance then it all adds up to one. A door is removed from the equation, the door that you didn't even choose, how can you say each door still has a 33% chance to freedom? 33% and 33% can not equal one, this means that one part of the equation is missing, so you must do something to make it equal one. The only obvious ways I see are to increase the odds 50%/50% or add that 33% back in there, thinking you can outrun the dogs.

This really, really, hurts right now, and I hope I made some sense, and I'm not 100% sure I'm right and open for discussion. I'm going to go put my head in ice now.
 
First thing's first. To clarify the coins riddle, there are nine coins in total and one of those nine is counterfeit, weighing slightly more. There are NOT ten coins. That being said, people have come close to the correct answer but have not stated it definitely. The answer should be the LEAST number of tries it would take to DEFINITELY know the counterfeit, without relying on choosing a lucky grouping that leaves the counterfeit out and balances perfectly. I'll leave it at that for a while.

Secondly, for those of you who are still arguing the doors riddle, all my respect and sympathy goes to you. You WILL need the ice for your head because it is not an easy riddle to solve. The problem with it is that it would be entirely different if the guard had eliminated the door before you made your first choice. Then it would be 50/50. In this case, though, it is 33% for the door you choose and 67% for the other door (it takes on the probability of the eliminated door. (It does have to add up to 100%, so VG30DETT is partially right.)
 
CONCLUSIVE PROOF OF THE "DOORS" RIDDLE:

Because this riddle is so hard to explain, I thought that conclusive evidence of Famine's answer would be nice. Imagine this:

You think of a number from one to three inclusive. You ask a friend to guess it. Your friend does so. You then eliminate a number that is A) not the right number and B) not the number your friend initially chose (so if you picked 3 and your friend guessed 1, you must eliminate 2). You then give your friend the opportunity to switch or to stick with his/her number. Does it matter what number you eliminated? The answer is no, because you will never eliminate your friend's guess. This is the key to the riddle. In effect, it is like you never eliminated a number. Now if you were to do this, say, 1,000 times and record the results each time, you would see that your friend has a better chance if he switches numbers than not.

Don't have the time? Don't have any friends? That's okay, because I wrote a program to do it for you. Here is the actual code:

RANDOMIZE TIMER
FOR x = 1 TO 1000
door = INT(RND * 3) + 1
choice = INT(RND * 3) + 1
elim:
eliminate = INT(RND * 3) + 1
IF eliminate = door OR eliminate = choice THEN GOTO elim
stay = INT(RND * 2)
IF stay = 0 AND choice = door THEN sright = sright + 1
IF stay = 0 AND choice <> door THEN swrong = swrong + 1
IF stay = 1 AND choice = door THEN cwrong = cwrong + 1
IF stay = 1 AND choice <> door THEN cright = cright + 1
NEXT x
CLS
PRINT "Stayed, Right:", sright
PRINT "Stayed, Wrong:", swrong
PRINT "Changed, Right:", cright
PRINT "Change, Wrong:", cwrong

It's in the QBASIC programming language, but since most people won't me able to understand it, here is what it does:

1000 times, it chooses a number randomly. Then it makes a guess randomly. Then it eliminated a number that is not the guess and not the correct number. Then it randomly chooses to stay or to switch. It tallies the number of times it is right and the number of times it is wrong. The results:

Stayed, Right: 160
Stayed, Wrong: 336
Changed, Right: 335
Change, Wrong: 169

Repeated several times, the results are very similar. The computer stayed 496 times and was right 160 of those times, or 32.3% of the time. It changed 504 times and was right 335 of those times, or 66.5% of the time. Staying was the right choice 160 + 169 = 329 times out of 1,000, or 32.9%. Changing, therefore, was the right choice 67.1% of the time. Those results are pretty indisputable.

By the way, the program could be changed so that the door was eliminated before the initial choice, like so:

RANDOMIZE TIMER
FOR x = 1 TO 1000
door = INT(RND * 3) + 1
elim:
eliminate = INT(RND * 3) + 1
IF eliminate = door THEN GOTO elim
chose:
choice = INT(RND * 3) + 1
IF choice = eliminate THEN GOTO chose
stay = INT(RND * 2)
IF stay = 0 AND choice = door THEN sright = sright + 1
IF stay = 0 AND choice <> door THEN swrong = swrong + 1
IF stay = 1 AND choice = door THEN cwrong = cwrong + 1
IF stay = 1 AND choice <> door THEN cright = cright + 1
NEXT x
CLS
PRINT "Stayed, Right:", sright
PRINT "Stayed, Wrong:", swrong
PRINT "Changed, Right:", cright
PRINT "Change, Wrong:", cwrong

(Notice that elim is before chose. Also notice that choice restricts itself now to only the two door remaining after elimination.)

The results, as you may be able to predict already, are similar to these:

Stayed, Right: 238
Stayed, Wrong: 243
Changed, Right: 263
Change, Wrong: 256

Draw your own conclusion...;)
 
Originally posted by ZZII
First thing's first. To clarify the coins riddle, there are nine coins in total and one of those nine is counterfeit, weighing slightly more. There are NOT ten coins. That being said, people have come close to the correct answer but have not stated it definitely. The answer should be the LEAST number of tries it would take to DEFINITELY know the counterfeit, without relying on choosing a lucky grouping that leaves the counterfeit out and balances perfectly. I'll leave it at that for a while.

Secondly, for those of you who are still arguing the doors riddle, all my respect and sympathy goes to you. You WILL need the ice for your head because it is not an easy riddle to solve. The problem with it is that it would be entirely different if the guard had eliminated the door before you made your first choice. Then it would be 50/50. In this case, though, it is 33% for the door you choose and 67% for the other door (it takes on the probability of the eliminated door. (It does have to add up to 100%, so VG30DETT is partially right.)

I get you now but explain to me WHY the **** the correct/right door takes on teh chance of the eliminated door. If you can explain that Ill accept defeat.
For the coins its 1-3 weighings:
Select 8 coins randomly and divide them into 4s, weigh the 2 groups. If they are the same you have found the counterfeit coin straight off, if one is heavier divide into pairs and take the heavier pair. Then pick the heavier coin out of the remaining 2.
 
About the doors.

Why dose the door you do NOT pick get the extra 33%?
If door 2 is elimated, then that 33% should be split in half, Giving each door 50%. Of coarse this is assuming you do not know the correct door. If you do know the correct door, then it is no longer probability.

So if you walk down a hallway, pick door 3. Then the doorman, or guard gets eaten by the dogs behind door 2, Why would changing your answer help you, or hurt you? Instead your your door (door 3) being 33% it just moved up to 50%.

The left over 33% from door 2 should not all go towards the right answer.

Simply because you picked a door, dose not make it wrong, or make the chances of probability not change. About the 10 doors, If 8 of them are wrong, that dose not make 1 of thoes doors 90% and one of them 10%. They would once again, be change to 50%, this is of coarse if you do NOT know the correct door. You are solving this riddle as you know that door 1 is correct and so you should give all the "left over" probability to that door.

If you think it dose, then go play the lottery. You know something about probability that no one else knows.
 
Your case has been argued several times in this post already. It dose make sense if the door was eliminated before you make your choice. Then thoes doors that remain would split the probability. The fact that you make a choice before the guard eliminates a door, though, dose change the riddle. Consider the case with 10 doors. The guard eliminates all but two. One must be the one you chose, regardless fo whether it is the correct one. The other door the guard leaves, in the 9 out of ten scenarios where the door you chose was not correct, must be the door that leads to the escape. Dose that make moer sense?
 
ok ZZII, you were the first one to explain it clearly. I can understand how that would work with the 10 doors. I guess I was kinda right! :rolleyes: That was a really tough puzzle to figure out. :p
 
I understood you the first time but your explaination is irrelevant because the guard doesnt know you've picked the left door. What if you'd picked the middle door obviously then you'd change because you're wrong. In taht situatiuon the 2 doors have the same probability because they both get the middle doors probability. What Im saying is it only works because its set up so you didnt choose the middle door(?) when in reality this would not always be the case.
 
Originally posted by Crayola
I understood you the first time but your explaination is irrelevant because the guard doesnt know you've picked the left door. What if you'd picked the middle door obviously then you'd change because you're wrong. In taht situatiuon the 2 doors have the same probability because they both get the middle doors probability. What Im saying is it only works because its set up so you didnt choose the middle door(?) when in reality this would not always be the case.

Well duh! It's a maths puzzle, not a situation extracted from the real-life memories of Camp X-ray prisoners and guards...

In any random choice, where you have already definitively picked one answer, and all but one remaining one are removed, you are always 50% (33% compared to 66%) better off changing your answer.

This was demonstrated in an American quiz show, where contestants had to pick the door (one of three) behind which the prize was. The host would remove one wrong door and give the contestant the choice of sticking or swapping. Amazingly, those who swapped were 50% more likely to win than those who stuck.
 
Originally posted by Famine
Well duh! It's a maths puzzle, not a situation extracted from the real-life memories of Camp X-ray prisoners and guards...

In any random choice, where you have already definitively picked one answer, and all but one remaining one are removed, you are always 50% (33% compared to 66%) better off changing your answer.

This was demonstrated in an American quiz show, where contestants had to pick the door (one of three) behind which the prize was. The host would remove one wrong door and give the contestant the choice of sticking or swapping. Amazingly, those who swapped were 50% more likely to win than those who stuck.

HAH! I finally got it visualising that gameshow. Thanks Famine. I heard about this problem a long time ago, never got it but now it all makes sense. It's so easy. I didn't care too much either when I saw this topic but now I just popped in and all pieces comes together.

Thanks suga!
 
I'm changing my answer to the original question (finally got around to reading it again and found that I read it wrong). the answer is B.
 
Originally posted by Wesley0913
ok ZZII, you were the first one to explain it clearly. I can understand how that would work with the 10 doors. I guess I was kinda right! :rolleyes: That was a really tough puzzle to figure out. :p

Yes, it is. :)

Originally posted by Jpec07
I'm changing my answer to the original question (finally got around to reading it again and found that I read it wrong). the answer is B.

Wow, I completely forgot about the lotto numbers with all this controversy over the doors riddle. Yes, B is the correct answer. (explaination somewhere back in the mess of posts)

As for the coins, the answer is two. No matter what happens, you can find the counterfeit coin in AT MOST two tries. How? Well several people are on the right track, leaving a coin off the balance. But nobody who has posted has got it exactly. Here is the answer:

Make three groups of three coins. Weigh two groups against each other. If one is heavier, it contains the counterfeit. If they are equal, the group not weighed contains the counterfeir. Next, weigh two coins from the group containing the counterfeit against each other. If one is heavier, it is the counterfeit. If they are ever, the coin not weighed is the counterfeit.

Yes, you may get lucky by weighing two groups of four and having them come up even, but it is not guaranteed. Two balances are needed to definately find the counterfeit.
 
Originally posted by Famine
Well duh! It's a maths puzzle, not a situation extracted from the real-life memories of Camp X-ray prisoners and guards...

In any random choice, where you have already definitively picked one answer, and all but one remaining one are removed, you are always 50% (33% compared to 66%) better off changing your answer.

This was demonstrated in an American quiz show, where contestants had to pick the door (one of three) behind which the prize was. The host would remove one wrong door and give the contestant the choice of sticking or swapping. Amazingly, those who swapped were 50% more likely to win than those who stuck.
Lol, Well at camp x-ray here a lot more doors and a lot less escaping... oh **** I've let the secret out... oh **** I definately shouldnt have told them it was a secret.
I understood you but im still trying to clarify it only works because he eliminates the door you didnt choose and if you werent to choose any then it would be fifty-fifty.
 
I would say i beleive famine in the doors problem. And ZZII just re enforced my belief. I cant say how or why but i understand and belive famine. ZZII proved well and truley that famine is right and the bit where he said you never eliminate the persons choice. This clicked in my brain and now i fully understand the problem but am unable to explain the why i understand it. If that makes sense
 
Yeah. I can see how the doors thing would work in that setup. I was thinking if you changed what door wasn't the correct one.

*in hippie voice* Think outside the box mannnnnn!
 
Originally posted by Small_Fryz
I would say i beleive famine in the doors problem. And ZZII just re enforced my belief. I cant say how or why but i understand and belive famine. ZZII proved well and truley that famine is right and the bit where he said you never eliminate the persons choice. This clicked in my brain and now i fully understand the problem but am unable to explain the why i understand it. If that makes sense

That makes perfect sense. It's why this riddle is so frustrating...even if you get it, you can't explain WHY. ;)
 
Go back and reread the question, Burnout. the first guy has a 1/1,000,000 chance at winning each week, and just because he didn't win the wek before doesn't mean that his chances are any better. However, the other guy has a 52/1,000,000 chance of winning because he plays all those numbers at once. I think 52/x is greater than 1/x.
 
Originally posted by Jpec07
Go back and reread the question, Burnout. the first guy has a 1/1,000,000 chance at winning each week, and just because he didn't win the wek before doesn't mean that his chances are any better. However, the other guy has a 52/1,000,000 chance of winning because he plays all those numbers at once. I think 52/x is greater than 1/x.

Everything you said is correct, but let me clear up the math a little:

For the person who picks 52 sets of numbers in one shot, you have it correct: 52(1/1,000,000) or whatever the odds are, so we'll say 52/X.

For the other person, who bets on one set of number 52 seperate times, it is a little more complicated. The odds that they get it on any of the tries is 1/X. However, you can't just add them together 52 times. If this was the case, then I could say that since I have a 1/2 chance of guessing the flip of a coin correctly, if I guess twice I will always be right once at least once. The correct way to combine them is much more involved and leads to a lower probability:

probability of winning the first time
+
(probability of losing the first time)*(probability of winning the second time)
+
(probabiity of losing the first two)*(probability of winning the third)
.........

By factoring out the probability of winning the first time, you would get:
1/X * (1 + (X-1)/X + ((X-1)/X)^2.......)

which must be less than 52/X

...i think:confused: :lol: ;)
 
Back