what's the sum of the numbers from 1 to 1,000?

add 1000 + 0, 999 + 1, 998 + 2, etc.. you get 1000 every time right down to 499 + 501.
that is 500 sets of 1000, so 500 * 1000 = 500,000 plus the final 500 right in the middle
that doesn't have a partner. 500,500
aha:!!!!!
a plane crashes on the border between canada and the u.s.. where should we bury the
survivors?
you don't bury survivors.

reverse a string - word by word aha:!!

a typical programming interview question is "reverse a string, in place". if

you understand pointers, the solution is simple. even if you don't, it can be
accomplished using array indices. i usually ask candidates this question first,
so they get the algorithm in their head. then i play dirty by asking them to
reverse the string word by word, in place. for example if our string is "the
house is blue", the return value would be "blue is house the". the words are
reversed, but the letters are still in order (within the word). solution:
reverse a string - word by word

problem: reverse "the house is blue", the answer should be "blue is house
the". the words are reversed, but the letters are still in order (within the
word).

solution: solving the initial problem of just reversing a string can either be a
huge help or a frustrating hinderance. most likely the first attempt will be to
solve it the same way, by swapping letters at the front of the string with
letters at the back, and then adding some logic to keep the words in order.
this attempt will lead to confusion pretty quickly.

for example, if we start by figuring out that "the" is 3 letters long and then
try to put the "t" from "the" where the "l" from "blue" is, we encounter a
problem. where do we put the "l" from "blue"? hmm... well we could have
also figured out how long "blue" was and that would tell us where to put the
"l" at... but the "e" from "blue" needs to go into the space after "the". argh.
its getting quite confusing. in fact, i would be delighted to even see a solution
to this problem using this attack method. i don't think its impossible, but i
think it is so complex that it's not worth pursuing.

here's a hint. remember before when we just reversed "the house is blue"?
what happened?

initial: the house is blue

reverse: eulb si esuoh eht

look at the result for a minute. notice anything? if you still don't see it, try
this.

initial: the house is blue

reverse: eulb si esuoh eht
wanted : blue is house the

the solution can be attained by first reversing the string normally, and then
just reversing each word.

100 doors in a row aha:!

problem: you have 100 doors in a row that are all initially closed. you make 100 passes
by the doors starting with the first door every time. the first time through you visit every
door and toggle the door (if the door is closed, you open it, if its open, you close it). the
second time you only visit every 2nd door (door #2, #4, #6). the third time, every 3rd
door (door #3, #6, #9), etc, until you only visit the 100th door.
question: what state are the doors in after the last pass? which are open which are closed?

solution: 100 doors in a row

problem: you have 100 doors in a row that are all initially closed. you make
100 passes by the doors starting with the first door every time. the first time
through you visit every door and toggle the door (if the door is closed, you
open it, if its open, you close it). the second time you only visit every 2nd
door (door #2, #4, #6). the third time, every 3rd door (door #3, #6, #9),
etc, until you only visit the 100th door.
for example, after the first pass every door is open. on the second pass you
only visit the even doors (2,4,6,8...) so now the even doors are closed and
the odd ones are opened. the third time through you will close door 3
(opened from the first pass), open door 6 (closed from the second pass),
etc..

question: what state are the doors in after the last pass? which are open
which are closed?

solution: you can figure out that for any given door, say door #42, you will
visit it for every divisor it has. so 42 has 1 & 42, 2 & 21, 3 & 14, 6 & 7. so on
pass 1 i will open the door, pass 2 i will close it, pass 3 open, pass 6 close,
pass 7 open, pass 14 close, pass 21 open, pass 42 close. for every pair of
divisors the door will just end up back in its initial state. so you might think
that every door will end up closed? well what about door #9. 9 has the
divisors 1 & 9, 3 & 3. but 3 is repeated because 9 is a perfect square, so you
will only visit door #9, on pass 1, 3, and 9... leaving it open at the end. only
perfect square doors will be open at the end.

red marbles, blue marbles aha:!

problem: you have two jars, 50 red marbles, 50 blue marbles. you need to
place all the marbles into the jars such that when you blindly pick one marble
out of one jar, you maximize the chances that it will be red. (when picking,
you'll first randomly pick a jar, and then randomly pick a marble out of that
jar) you can arrange the marbles however you like, but each marble must be
in a jar. solution: red marbles, blue marbles

problem: you have two jars, 50 red marbles, 50 blue marbles. you need to
place all the marbles into the jars such that when you blindly pick one marble
out of one jar, you maximize the chances that it will be red. (when picking,
you'll first randomly pick a jar, and then randomly pick a marble out of that
jar) you can arrange the marbles however you like, but each marble must be
in a jar.

solution: chance! chance is easy if you know how to do the formula. we know
that we have two choices to make. first we'll pick a jar, and each jar will
have a 1/2 chance of being picked. then we'll pick a marble, and depending
how we stack the marbles, we'll have a (# of red marbles in jar)/(# of total
marbles in jar) chance of getting a red one.

for example, say we put all the red marbles into jar A and all the blue ones
into jar B. then our chances for picking a red one are:

1/2 chance we pick jar A * 50/50 chance we pick a red marble

1/2 chance we pick jar B * 0/50 chance we pick a red marble

do the math and you get 1/2 chance for a red marble from jar A and a 0/2
chance for a red marble from jar B. add 'em up and you get the result = 1/2
chance for picking a red marble.

think about it for awhile and see if you can figure out the right combination.
we had a 50/50 (guaranteed) chance in picking a red marble from jar A, but
we didn't have to have 50 red marbles in there to guarantee those fantastic
odds, did we? we could've just left 1 red marble in there and the odds are
still 1/1. then we can take all those other marbles and throw them in jar B to
help the odds out there.

let's look at those chances:

1/2 we pick jar A * 1/1 we pick a red marble

1/2 we pick jar B * 49/99 we pick a red marble

do the math and add them up to get 1/2 + 49/198 = 148/198, which is
almost 3/4.
we can prove these are the best odds in a somewhat non-formal way as
follows. our goal is to maximize the odds of picking a red marble. therefore
we can subdivide this goal into maximizing the odds of picking a red marble
in jar A and maximizing the odds of picking a red marble in jar B. if we do
that, then we will have achieved our goal. it is true that by placing more red
marbles into a jar we will increase the chances of picking a red marble. it is
also true that by reducing the number of blue marbles in a jar we will
increase the odds also. we've maximized the odds in jar A since 1/1 is the
maximum odds by reducing the number of blue marbles to 0 (the minimum).
we've also maximized the number of red marbles in jar B. if we added any
more red marbles to jar B we would have to take them out of jar A which
reduce the odds there to 0 (very bad). if we took any more blue ones out of
jar B we would have to put them in jar A which reduce the odds there by
50% (very bad).

it wasn't really a good proof, but QED anyway :-P

switches puzzle aha:!

This one was recently given on Car Talk, and it's been making the rounds right now
also...
The warden meets with 23 new prisoners when they arrive. He tells them, "You may
meet today and plan a strategy. But after today, you will be in isolated cells and will have
no communication with one another.
"In the prison is a switch room, which contains two light switches labeled A and B, each
of which can be in either the 'on' or the 'off' position. I am not telling you their present
positions. The switches are not connected to anything.
"After today, from time to time whenever I feel so inclined, I will select one prisoner at
random and escort him to the switch room. This prisoner will select one of the two
switches and reverse its position. He must move one, but only one of the switches. He
can't move both but he can't move none either. Then he'll be led back to his cell.
"No one else will enter the switch room until I lead the next prisoner there, and he'll be
instructed to do the same thing. I'm going to choose prisoners at random. I may choose
the same guy three times in a row, or I may jump around and come back.
"But, given enough time, everyone will eventually visit the switch room as many times as
everyone else. At any time anyone of you may declare to me, 'We have all visited the
switch room.' and be 100% sure.
"If it is true, then you will all be set free. If it is false, and somebody has not yet visited
the switch room, you will be fed to the alligators."
What is the strategy they come up with so that they can be free?

cars on the road aha:!

if the probability of observing a car in 20 minutes on a highway is 609/625,

what is the probability of observing a car in 5 minutes (assuming constant
default probability)?

thanks to joel w. for this one

railroad bridge aha:!!

i got some new brain teasers and more classic programming puzzles for
everyone!

a man needs to go through a train tunnel. he starts through the tunnel and
when he gets 1/4 the way through the tunnel, he hears the train whistle
behind him. you don't know how far away the train is, or how fast it is
going, (or how fast he is going). all you know is that

1. if the man turns around and runs back the way he came, he will just
barely make it out of the tunnel alive before the train hits him.
2. if the man keeps running through the tunnel, he will also just barely
make it out of the tunnel alive before the train hits him.

assume the man runs the same speed whether he goes back to the start or
continues on through the tunnel. also assume that he accelerates to his top
speed instantaneously. assume the train misses him by an infintisimal
amount and all those other reasonable assumptions that go along with
puzzles like this so that some wanker doesn't say the problem isn't well
defined.

how fast is the train going compared to the man?

chameleons aha:!
"at one point, a remote island's population of chameleons was divided as
follows:

• 13 red chameleons
• 15 green chameleons
• 17 blue chameleons

each time two different colored chameleons would meet, they would change
their color to the third one. (i.e.. If green meets red, they both change their
color to blue.) is it ever possible for all chameleons to become the same
color? why or why not?"

another one from levik

crazy guy on the airplane aha:!

"a line of 100 airline passengers is waiting to board a plane. they each hold a
ticket to one of the 100 seats on that flight. (for convenience, let's say that
the nth passenger in line has a ticket for the seat number n.)

unfortunately, the first person in line is crazy, and will ignore the seat
number on their ticket, picking a random seat to occupy. all of the other
passengers are quite normal, and will go to their proper seat unless it is
already occupied. if it is occupied, they will then find a free seat to sit in, at
random.

what is the probability that the last (100th) person to board the plane will sit
in their proper seat (#100)?"

wanna play? aha:!

i offer to play a card game with you using a normal deck of 52 cards. the
rules of the game are that we will turn over two cards at a time. if the cards
are both black, they go into my pile. if they are both red, they go into your
pile. if there is one red and one black, they go into the discard pile. we
repeat the two card flipping until we've gone through all 52 cards. whoever
has more cards in their pile at the end wins. i win if there is a tie. if you
win, i pay you a dollar. how much would you pay to play this game?

solution: wanna play?

I wouldn't give you a penny for that game. Here's why.

Let's say all the color pairs are matched. That means 13 red pairs, 13 black
pairs, you win.

Now let's say there is 12 red pairs in the deck, we pick them all from the top
of the deck. With the 28 remaining card, there will always be exactly 12
black pairs and two mixed pairs. Why? Two red cards in the stack, they can't
be in a pair, otherwise you would get a total of 13 pairs, and I would too
since all the black cards would be together. So the two red cards are
matched with one black card each, leaving 24 black cards together, I get the
same number of pairs as you do and you win.

Same math with 11 red pairs. That means four red cards in mixed pairs,
hence 22 black cards left together, or 11 pairs.

No matter how mixed up the cards are, it's a tie, and you win. Hence, I
wouldn't give you a penny.

smart cookie aha:?

did you ever wonder how they make those pillsbury cookie dough rolls with
the intricate faces inside them? scroll about halfway down the page here and
notice the intricate design they have somehow injected into their cookie
rolls? if you examine the roll closely there is no seam between the normal
dough and the colored shape, but somehow they get that inside the roll. i
emailed them asking them how they do it and they told me it was "doughboy
magic".

solution: smart cookie

I don't know if this is right or not, but it sounds pretty darn close to the best
idea I've heard yet.

From Brad Siemssen:

"I assume these are made the same way designs are put in the center of
taffy candies.

The designs are made by taking advantage of the fact that the dough is soft
and pliable.

They take the dough and make the design at a large diameter and thickness.
The diameter could a foot or more. At this scale the design does not have to
be very precise. Extrusion methods like the playdough presses would work.

They then take the giant dough cylinder with the rough design and then roll
it into a "snake" just like kids do with balls of playdoh. As the "snake" gets
longer the design in the center gets uniformly smaller. The "snake" is then
cut into pieces and put in packaging.

To get a more precise designs simply requires increasing the starting

diameter. The engineering difficulty then lies in rolling the snake out.
An alternative to rolling would be to use a press to extrude the dough using a
funnel shaped press. The engineering difficulty with a press would be to
shape the press, and exert uniform pressure so the dough coming out is a
shrunken version of the original."

XOR using NAND gates aha:!

I can't remember if I saw this on the discussion board or somewhere else...

Create an XOR gate using only NAND gates. (BTW, mostly all circuit
problems assume you have two inputs plus 0 and 1 also as inputs).

x implies y aha:!!

Part I

(you can't use paper, you have to figure it out in your head)

i have a black triangle, a white triangle, a black circle and a white circle. if i
gave you a shape (triangle or circle) and a color (black or white), the
"frobby" items would be those that had either the shape or the color, but not
both. that is, in order to be frobby, the item must be of the specified color
OR the specified shape, but not both the specified shape AND the specified
color. i'm thinking of a shape and a color in my head and i tell you that the
white triangle is frobby. can you tell me the "frobbiness" of the other items?

Part II

there are four cards which have a letter on one side and a number on the
other side. i lay them out and the cards appear as 2 5 F E. the rule is that a
card with an odd number on one side must have a vowel on the other. what
is the minimum number of cards you should turn over to prove the rule is
true (and which cards would they be)?
technical riddles aha:!!

You have an empty room, and a group of people waiting outside the room. At
each step, you may either get one person into the room, or get one out. Can
you make subsequent steps, so that every possible combination of people is
achieved exactly once?

black dots aha:!

moving circles aha:!

line mirage aha:!

even harder coin problem aha:!

some of you may have easily solved the pill weighing problem posed here. if
so you are going to love this problem.

it is similar but much more difficult.

from my buddy tom:

> ok, here's a tough one (i thought). there are no "aha!"

tricks - it
> requires straightforward deductive-reasoning.
>
> you have 12 coins. one of them is counterfeit. all the
good coins weigh
> the same, while the counterfeit one weights either more
or less than a
> good coin.
>
> your task is to find the counterfeit coin using a
balance-scale in 3
> weighs. moreover, you want to say whether the coin
weighs more or less
> than is should and, and this is the real kicker, your
weighs must be
> non-adaptive.
>
> that is, your choice of what to put on the balance for
your
> second weigh cannot depend on the outcome of the first
weigh and your
> decision about what to weigh for round 3 cannot depend on
what happened on
> either your first or second weigh.
>
> for example, you can't say something like "take coin #1
and coin #2 and
> weigh them. if they balance, then take coins 3,4,5 and
weight them
> against 6,7,8...if 1 and 2 don't balance, then weigh #1
vs #12..." you
> have to say something like:
>
> round #1: do this
> round #2: do this
> round #3: do this
>
> if the results are left tilt, balanced, and left tilt,
respectively, then
> coin #11 is heavier than it should be.
>
> this problem is solveable...it took me about 1-2 hours of
working on it to
> get it. i think even finding the counterfeit using an
adaptive solution
> is tough. then non-adaptive constraint makes it quite
hard and having to
> find whether it's heavier and lighter is cruel and
unusual riddling ;-)
>
> have fun...

box o' numbers aha:!

arrange the numbers 1 to 8 in the grid below such that adjacent numbers are
not in adjacent boxes (horizontally, vertically, or diagonally).

___

| 1 |

=============

| 6 | 4 | 3 |

=============

| 2 | 7 | 5 |

=============

| 8 |

=====

the arrangement above, for example, is wrong because 3 & 4, 4 & 5, 6 & 7,
and 7 & 8 are adjacent.

solution: box o' numbers

The key (as I see it) is putting the 1 & 8 in the centre spots - which is
required, because those spots both border all but one of the other spots, and
1 & 8 are the only numbers that are only adjacent to one number.

From there, the 2 & 7 are forced, then you have your choice of the next
number, which then forces the rest. Really, though there are only two valid
solutions, and they are mirror images of each other.

even harder coin problem aha:!

some of you may have easily solved the pill weighing problem posed here. if
so you are going to love this problem.

it is similar but much more difficult.

from my buddy tom:

> ok, here's a tough one (i thought). there are no "aha!"
tricks - it
> requires straightforward deductive-reasoning.
>
> you have 12 coins. one of them is counterfeit. all the
good coins weigh
> the same, while the counterfeit one weights either more
or less than a
> good coin.
>
> your task is to find the counterfeit coin using a
balance-scale in 3
> weighs. moreover, you want to say whether the coin
weighs more or less
> than is should and, and this is the real kicker, your
weighs must be
> non-adaptive.
>
> that is, your choice of what to put on the balance for
your
> second weigh cannot depend on the outcome of the first
weigh and your
> decision about what to weigh for round 3 cannot depend on
what happened on
> either your first or second weigh.
>
> for example, you can't say something like "take coin #1
and coin #2 and
> weigh them. if they balance, then take coins 3,4,5 and
weight them
> against 6,7,8...if 1 and 2 don't balance, then weigh #1
vs #12..." you
> have to say something like:
>
> round #1: do this
> round #2: do this
> round #3: do this
>
> if the results are left tilt, balanced, and left tilt,
respectively, then
> coin #11 is heavier than it should be.
>
> this problem is solveable...it took me about 1-2 hours of
working on it to
> get it. i think even finding the counterfeit using an
adaptive solution
> is tough. then non-adaptive constraint makes it quite
hard and having to
> find whether it's heavier and lighter is cruel and
unusual riddling ;-)
>
> have fun...

mensa aha:!

if "24 H in a D" means "24 hours in a day", what does "26 L of the A"
mean? what about the other 32 puzzles here on my mensa forward
reproduction page?
Wednesday, August 22, 2001

i received this as an .xls attachment in the mail. i would just post the excel
file, but i don't know if it has a virus or not, so i'm just going to copy the
text here:

0 24 H in a D
1 26 L of the A

2 7 D of the W

3 7 W of the W

4 12 S of the Z

5 66 B of the B

6 52 C in a P (WJs)

7 13 S in the USF

8 18 H on a G C

9 39 B of the O T

10 5 T on a F

11 90 D in a R A

12 3 B M (S H T R)

13 32 is the T in D F at which W F

14 15 P in a R T

15 3 W on a T

16 100 C in a R

17 11 P in a F (S) T

18 12 M in a Y

19 13=UFS

20 8 T on a O

21 29 D in F in a L Y

22 27 B in the N T

23 365 D in a Y

24 13 L in a B D

25 52 W in a Y

26 9 L of a C

27 60 M in a H

28 23 P of C in the H B

29 64 S on a C B

30 9 P in S A

31 6 B to an O in C

32 1000 Y in a M
 33 15 M on a D M C

you have to figure out what they all mean. for example, the first one is "24
hours in a day".

i'll post the answers later (post your answers on the discussion board if you
feel the need).

hen aha:!!!

if a hen and a half lay an egg and a half in a day and a half, how many hens
does it take to lay six eggs in six days?

solution: hen

if 1.5 hens lay 1.5 eggs in 1.5 days (or 36 hours) then: 1 hen lays 1 egg in
1,5 days or 4 eggs in six days thus 1.5 hens lay 6 eggs in 6 days

duel aha:!!

you find yourself in a duel with two other gunmen. you shoot with 33%
accuracy, and the other two shoot with 100% and 50% accuracy,
respectively. the rules of the duel are one shot per-person per-round. the
shooting order is from worst shooter to best shooter, so you go first, the
50% guy goes second, and the 100% guy goes third.

where or who should you shoot at in round 1?

solution: duel

You have 3 options to consider:

(1) Shoot at 50% guy. Death comes with: 72% likilhood (2) Shoot at the
100% guy. Death comes with: 63.64% likihood (3) Shoot into the air
(purposely miss). Death comew with 58.33% likihood.

So, you shoot into the air and hope for the best.

Calculations:

The p in the probabilities below represents your probability of dying at some

point if you MISS either guy in the first round. It'll be filled in later.

(1) if you shoot at the 50% guy and get him, you're guaranteed to get shot
the next round.

prob of dying at some point by aiming at the 50% guy:

(=) (33.3%)(100%)+(66.67%)*(p)

(=) (33.3%)+(66.67%)(p)

(2) if you shoot at the 100% guy and get him, you're left with this geometric
sum of probability of getting shot by the 50% guy at some point in the
future...it converges.

(=) [(50%)] +

[(50%)(66.67%)(50%)] +

[(50%)(66.67%)(50%)(66.67%)(50%)] + ...

(=) 50% * { 1 + (1/3) + (1/3)^2 + ... }

(=) 50% * {3/2)

(=) 75%
so, your prob of getting shot if you shoot at the 100% guy is:

(=) (33.3%)(75%)+(66.67%)(p)

(=) (24.75%) + (66.67%)(p)

still crummy, but it dominates the alternative, for positive p.

(3) What happens if you miss? ie, what's (p)?

if you miss, the second guy has a choice to make...does he shoot at you, or
the other guy? his options:

(1) Shoot you: if he gets you, he's guaranteed to die on the next shot.

if he misses, he has q chance of dying (which we'll get later).

(=) (50%)(100%)+(50%)(q)

(=) (50%)+(0.5)(q)

(2) Shoot at the 100% dude:

If he gets the 100% guy, there some infinite sum representing his probability
of getting shot by you:

[(33.33%)] +

[(66.67%)(50%)(33.33%)] +

[(66.67%)(50%)(66.67%)(50%)(33.33%)] + ...

(=) 33.33% * [ 1 + (1/3) + (1/3)^2 + ... ]

(=) 33.33% * [(3/2)]

(=) (50.0%)
So, he chances of getting shot if he shoots at the 100% guy are:

(=) (50%)*(50%) + (50%)(q)

(=) (25%)+(.5)q

No matter what q is, he'd prefer shooting the 100% guy to shooting you.

Now, what happens if *he* misses (the 100%) guy? ie, what's q? if he
misses, then the 100% guy has to make a decision:

(1) Shoot you:

he's guaranteed to get you, so his chances of dying are just 50%. (game
ends after the next round...the 50% gets only one shot)

prob of dying: 50%

(2) Shoot the 50% guy:

by the same logic, he'd have a 33.33% chance of death (getting shot by
you).

So, he prefers to shoot the 50% guy.

So, q is 100% (ie, if the 50% guy misses, the 100% guy shoots him
immediately). From that, we know the 50% guy's optimal move, if
everyone's around on his first shot. He'd prefer shooting at the 100% guy to
shooting at you or purposely missing.

Now, we need to get p, which is our probability of dying if we purposely miss

either guy (shoot into the air).

We know the 50% guy shoots at the 100% guy if we miss.

(1) With 50% probability, the 50% guy kills the 100% guy, resulting in a
shootout b/w us and the 50% guy.

(=) (50%) * [ (66.67%)*(50%) + (66.67%)*(50%)*(66.67%)(50%) + ...

(=) (50%) * [ (1/3) + (1/3)^2 + ... ]

(=) (50%) * [0.5]

(=) (25%)

(2) He misses him. Then, we get one shot at the 100% guy (after the 100%
guy shoots the 50% guy).

Our chances of death are:

(50%)*(66.67%) = (33.33%)

So, p is (25%)+(33.33%) = 58.3%

So, again, our three choices are

(1) Shoot at 50% guy. Death comes with:

(=) (33.3%)+(66.67%)(p)

(=) 72% likilhood

(2) Shoot at the 100% guy. Death comes with:

(=) (24.75%) + (66.67%)(p)

(=) 63.64% likihood

(3) Shoot into the air (purposely miss).

(=) p
(=) 58.33% likelihood.

vienna aha:!!

it's the middle ages, you're travelling across europe and you want to find the
way to vienna. you come to a crossroads, now there are two ways to go. at
the crossroads stand a knight and a knave. the knight answers every
question truthfully. the knave answers every question falsely. you don't know
which guy is which. how can you figure out which road leads to Vienna by
only asking one question?

solution: vienna

Many people try "what would the other man say if I asked him which way to
Vienna?" But for that to work, the knight would have to know that the other
man is a knave, and the knave would have to know that the other man is a
knight; it's not clear that either of these is a given.

To cope with that, you can ask "assuming the other man has the opposite
predilection regarding truth-telling from what you do, what would he say if I
asked him which way to Vienna?"

More interestingly, you can ask "if you were to ask yourself which way to
Vienna, what would you say?" But before risking that question, I'd want a
few more details about how the knave is wired up. What does it mean, to lie
to oneself? Is such a thing even possible, regardless of what psychoanalysts
might claim?

Furthermore, travelling through Europe in the middle ages, it's a bit of a rash
assumption that people you meet at a crossroad would understand English. If
you only have 1 question, you have to frame a question that is meaningful in
all locally spoken languages (like writing a program that is both legal Fortran
and legal C). Your question doesn't have to mean exactly the same thing in
each language, as long as each possible answer carries in itself an indication
of which language it is in.

Or you could try the universal language of mime. Say "Vienna" and open
your arms wide with a questioning look. But how would you mime "if you
were to ask the other man ..." (or "if you were to ask yourself ...") ?

oil mogul aha:!

you are an oil mogul considering the purchase of drilling rights to an as yet
unexplored tract of land.

the well's expected value to its current owners is uniformly distributed over
[$1..$100]. (i.e., a 1% chance it's worth each value b/w $1..$100, inclusive).

bcause you have greater economies of scale than the current owners, the
well will actually be worth 50% more to you than to them (but they don't
know this).

the catch: although you must bid on the well before drilling starts (and
hence, before the actual yield of the well is known), the current owner can
wait until *after* the well's actual value is ascertained before accepting your
bid or not.

what should you bid?

solution: oil mogul

This problem amounts to properly defining the expected value of the well to
you.

The following equation does it:

(1%) * [(1.5 - 1)] +

(1%) * [(3 - 2) + (1.50 - 2)] +

(1%) * [(4.5 - 3) + (3 - 3) + (1.5 - 3)] +

...

(1%) * [(150-100) + ... + (3-100) + (1.5-100)]

Each line represents your expected value from a bid of 1$, 2$, ..., 100$,
respectively.

eg, consider line 2 above. if you bid $2...

With 98% probability you won't win the contract, so your profit is 0. With 1%
probability, you will win something worth (150%*1) = 1.5, for which you
paid 2$ With 1% probability, you will something worth worth (150%*2) = 3,
for which you paid 2$

So, your goal is to maximize the following function of x, where x is your bid.

f(x) = 1% * Sum_{i = 1 to floor(x)}{ x - 1.5*i }

There's no benefit to non-integer bets, so re-write the maximization function

as :

ARGMAX(k) {1% * Sum_{i = 1 to k}{1.5*i - k}}

(=) ARGMAX(k) {Sum_{i=1 to k}{1.5*i - k}} /* 1% isn't a function of k or i,

so toss it */

(=) ARGMAX(k) {Sum_{i=1 to k}{1.5*i} - Sum_{i=1 to k}{k}} /* Split the

summation */

(=) ARGMAX(k) {(0.75)(K)(K+1) - K^2 }} /* Closed form for summations */

(=) ARGMAX(k) {(0.75)(k)-(0.25)(K^2)}} /* Algebra */

And that function is maximized at k = 1 and k = 2.

When choosing b/w $1 and $2, you should bid $1 because of time-value,
reinvestment risk, etc, of the extra dollar.

(ie, if you don't have to spend the extra $$ now, don't)

100 factorial aha:!

how many trailing zeroes are there in 100! (100 factorial)?

solution: 100 factorial

One per factor of 10, and one per factor of 5 (there are more than enough
2's to pair with the 5's), plus one per factor of ten squared (one occurrence)
and one per factor of 5 squared (three occurrences).

So if I'm counting correctly, that'd be 10 + 10 + 1 + 3== 24 zeroes.

Assuming the question meant *trailing* zeroes. It'd be much harder to also
count the intermingled zero digits in the entire expansion.

world series aha:!

you have $10,000 dollars to place a double-or-nothing bet on the Yankees in

the World Series (max 7 games, series is over once a team wins 4 games).

unfortunately, you can only bet on each individual game, not the series as a
whole. how much should you bet on each game, so that, if the yanks win the
whole series, you expect to get 20k, and if they lose, you expect 0?

basically, you know that there may be between 4 and 7 games, and you
need to decide on a strategy so that whenever the series is over, your final
outcome is the same as an overall double-or-nothing bet on the series.

solution: world series

this probably isn't the cleanest solution, but...

a dynamic-programming type solution is:

(1) Create a 5x5 matrix P.

So, P[i,j] holds your pile of money when the yanks have won i games and the
mets have won j games.

initialize P[4,j] := 20 for j from 0 to 3 initialize P[i,4] := 0 for i from 0 to 3

fill P in bottom-right to top left by averaging bottom and right adjacent cells:

P[i,j] := (P[i+1,j]+P[i,j+1]) / 2

(2) Make another 5x5 matrix, B, which represets your bet at any-time.

So, B[i,j] represents your bet when the yanks have won i games and the
Mets j games.

fill this top-left to bottom right by:

B[i,j] = P[i+1,j] - P[i,j]

(3) Look in matrix B for your bet at any time.

The final matricies are:

Pile-
Matrix

0.00 1.00 2.00 3.00 4.00

0 10.00 6.88 3.75 1.25 0.00

1 13.13 10.00 6.25 2.50 0.00

2 16.25 13.75 10.00 5.00 0.00

3 18.75 17.50 15.00 10.00 0.00

4 20.00 20.00 20.00 20.00

Bet-
Matrix

0 1 2 3 4

0 3.13 3.13 2.50 1.25

1 3.13 3.75 3.75 2.50

2 2.50 3.75 5.00 5.00

3 1.25 2.50 5.00 10.00

salary aha:!

three coworkers would like to know their average salary. how can they do it,
without disclosing their own salaries?

solution: salary

How about: Person A writes a number that is her salary plus a random
amount (AS + AR) and hands it to B, without showing C. B then adds his
salary plus a random amount (BS + BR) and passes to C (at each step, they
write on a new paper and don't show the 3rd person). C adds CS + CR and
passes to A. Now A subtracts her random number (AR), passes to B. B and C
each subtract their random number and pass. After C is done, he shows the
result and they divide by 3.

As has been noted already, there's no way to liar-proof the scheme.

It's also worth noting that once they know the average, any of the three
knows the sum of the other 2 salaries.

orbs aha:!
you have two identical crystal orbs. you need to figure out how high an orb
can fall from a 100 story building before it breaks. you know nothing about
the toughness of the orbs: they may be very fragile and break when dropped
from the first floor, or they may be so tough that dropping them from the
100th floor doesn't even harm them.

what is the largest number of orb-drops you would ever have to do in order
to find the right floor? (i.e. what's the most efficient way you could drop the
orbs to find your answer?)

you are allowed to break both orbs, provided that in doing so you uniquely
identify the correct floor.

solution: orbs

14.

Drop first orb from floors 14, 27, 39, 50, 60, 69, 77, 84, 90, 95, 99, 100...
(ie move up 14 then 13, then 12 floors, etc) until it breaks (or doesn't at
100). Call the first floor at which it breaks n and the previous tested floor n'.
Then try the intervening floors (n'+1 .. n'-1) with the other orb.

Worst case is if correct floor is 13,14,26,27, etc which require m drops with
the first orb and 14-m drops with the second.

nuggets aha:!

you can go to a fast food restaurant to buy chicken nuggets in 6-pack, 9-

pack or 20-packs. is there such a number N, such that for all numbers bigger
than or equal to N, you can buy that number of chicken nuggets?

solution: nuggest

Here's another way of looking at it:

 1 2 3

4 5 6

7 8 9

10 11 12

13 14 15

16 17 18

19 20 21

22 23 24

25 26 27

28 29 30

31 32 33

34 35 36

37 38 39

40 41 42

43 44 45

(and so on, to infinity)

You can get any number in the right column except 3 by adding 6s and 9s.
So cross out the entire right column except 3. You can also add 20 to any
crossed-out number and cross that number out. So cross out everything in
column two below and including 26. Finally, by the same logic you can add
20 to the crossed-out numbers in column 2 and thereby cross out everything
in column one below and including 46.

The largest number that's left over is 43. Incidentally, cross out 20 and 40
and your map is complete.

chinese emperor aha:!

A chinese emperor had to choose a new adviser amongst 3 sages, all of them
equally wise. He placed a problem to them: "To choose one of you, you'll
play a simple and fair game: In this sack there are 3 white balls and 2 black
balls. Each of you will be blindfolded and will pick one ball and place it on
your head. After that, the blindfolds will be removed and each one in turn will
try to guess the colour of the ball upon his head, by observation of the other
picked balls. However, beware. You may pass your turn in guessing, but if
you state a colour and fail, you're disqualified. This way I'll learn which one is
the most intelligent amongst you" The sages talked briefly to each other and
promptly refused: "Dear lord, it's of no use, since the game is not fair. The
last one of us to guess in the first round will know the answer." and the
sages promptly demonstrated this to the emperor, who was so amazed by
their wits that he appointed all 3 has his advisers. Could you demonstrated it
? NOTE: If the emperor had any wits at all he would have named them all
advisers in the first place... maybe spending reduction ? :)

painfully easy aha:!

i flip a penny and a dime and hide the result from you. "one of the coins
came up heads", i announce. what is the chance that the other coin also
came up heads?

solution: painfully easy

Assuming complete honesty on the part of the flipper, wouldn't the solution
be 33%?

There are four possible scenarios:

HH
TH
HT
TT
Obviously the TT possibility can be discounted because it does not result in
one of the two being heads.

This lees us with three possibilities, only one of which has the other coin also
being heads.

Therefore one third.

I think.

I usually get these wrong.

boys and girls aha:!

in a country in which people only want boys, every family continues to have
children until they have a boy. if they have a girl, they have another child. if
they have a boy, they stop. what is the proportion of boys to girls in the
country?

solution: boys and girls

Pretty simple. Half the couples have boys first, and stop. The rest have a girl.
Of those, half have a boy second, and so on.

So suppose there are N couples. There will be N boys. There will be an

"infinite" sum of girls equal to N/2 + N/4 + N/8 + ... As any college math
student knows, this sum adds up to N. Therefore, the proportion of boys to
girls will be pretty close to 1:1, with perhaps a few more boys than girls
because the sum isn't actually infinite.

Btw, hope you enjoyed your fishing trip, Michael!

calendar cubes aha:!!!!

a man has two cubes on his desk. every day he arranges both cubes so that
the front faces show the current day of the month. what numbers are on the
faces of the cubes to allow this?

note the aha factor

solution: calendar cubes

Hmm. I don't know why this would warrant four aha's, but we'll see...

First, to show all possible days, we'd need one of each of the ten digits. We'd
also need two 1s and two 2s to show 11 and 22. That's twelve numbers right
there. Two cubes, twelve faces, so every face is used. Quite elegant.

We know each cube will need a 1 and a 2. Let's put the 3 on one of them.
The 0 has to go on the other. We put 4, 5, and 6 on the 3-cube since we
need to show 04 05 06.

But now where do 7, 8, and 9 go? The 0-cube needs them to be on the 3-
cube, but it's full. I'm beginning to see why this gets four aha's.

Let's try this: clear both cubes, put the 0 on one of them. Now 1-9 have to
go on the other cube to show 01-09. We can't put a 0 on the other cube, too,
because that puts us over the 12-digit limit. It seems I have mathematical
proof that this cannot be done! What am I missing? It's not like you can turn
one of the other numbers sidewise to make another 0...

Heh.

You CAN make a 9 out of a 6, though. That frees up a digit. So you put 0, 1,
and 2 on both cubes. Put 3 on one of them. 4 and 5 can go on it too. Put 6,
7, and 8 on the other. Now you can show 01-31 with no problem, and even
00 and 32 if you're feeling weird.

I'd've given it two, maybe three aha's. :)

coin rolls aha:!

every night, i dump all the change in my pocket into a big bucket.

when I buy things, i never hand over coins. always bills. so i accumulate a lot
of coins. even if the purchase price is $1.01, and i have lots of coins in my
pocket, i pay $2 and take the 99c in change. all the more coins to dump in
my change bucket!

after about 10 years of this, i decide to roll all the coins into rolls. remember
that a quarter roll is $10, a dime roll is $5, nickels $2, pennies 50 cents. so I
go to the Banking Supply Store and buy empty paper rolls.

the Banking supply store, conveniently, sells assortment packs of coin rolls.
each assortment pack contains W quarter rolls, X dime rolls, Y nickel rolls,
and Z penny rolls.

the question: what is the optimum ratio of W to X to Y to Z to maximize the

probability that I will use up the assortment packs at the same rate, e.g.
without lots of leftover nickel tubes and stuff.

p.s. this problem should ideally be solved using Excel (but if you really like
doing these things by hand, be my guest).

submitted by joel

paul brinkley makes these assumptions which are all good assumptions to
make:
Assumption 1: The price of purchases made, modulo $1, is an even
distribution from 0 cents to 99 cents.
Assumption 2: The cashier will always give you the least number of coins
mathematically possible, and will always have enough of each type of coin to
do this. So you'll never get 99 pennies as change for a $1.01 purchase, for
example.
Assumption 3: Half dollars don't exist.

solution: coin rolls

"Brute force" approach:

I guess I'll begin with a few assumptions:

Assumption 1: The price of purchases made, modulo $1, is an even

distribution from 0 cents to 99 cents. Probably not true, but we can cover
that later.

Assumption 2: The cashier will always give you the least number of coins
mathematically possible, and will always have enough of each type of coin to
do this. So you'll never get 99 pennies as change for a $1.01 purchase, for
example.

Assumption 3: Half dollars don't exist.

Over the long haul, then, you'd get N sets of 0 cents, 1 cent, 2 cents, and so
on up to 99 cents. So let's consider one of each. How many of each coin
would you end up with?

Easy first: let's do quarters. 25-49 cents each gets you one. 50-74 each gets
you two. 75-99 each gets you three. That's (1+2+3)*25, or 150 quarters.

Dimes next. 10-19 gets you one. 20-24, two. (You'd get a quarter instead for
25 and up.) 35-44, one. 45-49, two. 60-69, 85-94, one. 70-74, 95-99, two.
So that's 4*(10+5*2), or 80 dimes.

Nickels. One for 5-9, 15-19, and that same pattern for +25, +50, +75, so it's
4*10 or 40 nickels. You'll never get two at a time, since a dime (at least) will
do.
Pennies. Let's cut to the chase: 1,2,3, and 4 cents gets 10 pennies in all, and
that pattern happens 20 times, so that's 200 pennies.

Let's double-check. 200 pennies, 40 nickels, 80 dimes, 150 quarters. That's

200*1 + 40*5 + 80*10 + 150*25 cents = 200+200+800+3750 = 4950
cents, which is the sum of all numbers from 0 to 99, so it checks.

15/8/4/20 would be the ratio then, IF coin rolls all hold the same number of
coins, but they don't. Quarter rolls hold 40 coins, dime rolls hold 50, nickel
rolls 40, penny rolls 50. So you need 5/4 as many quarter and nickel rolls.
The final ratio is 75/32/20/80. Seems like a lot of quarters and pennies,
except for the assumption that you'll tend to get them much more often than
nickels and dimes as change.

The numbers change slightly when you figure in things like frequency of
coins in circulation, the supply the cashier has at any one time, the actual
distribution of change values, and the cashier's inclination to give you two
dimes and a nickel instead of a quarter just because...

So what numbers do assortment packs really contain?

last ball aha:!

you have 20 blue balls and 14 red balls in a bag. you put your hand in and
remove 2 at a time. if they're of the same color, you add a blue ball to the
bag. if they're of different colors, you add a red ball to the bag. (assume you
have a big supply of blue & red balls for this purpose. note: when you take
the two balls out, you don't put them back in, so the number of balls in the
bag keeps decreasing). what will be the color of the last ball left in the bag?

once you tackle that, what if there are 20 blue balls and 13 red balls to start
with?

solution: last ball

You always take off Red Balls two by two !

So if you start with 14 Red Balls, you cannot have one single Red ball at the
end.

... so the last ball is blue.

But if you start with 13 Red Balls, as you take them off 2 by 2 (at one
moment or the other you'll do it !) you will arrive at a moment when you
have 1 red ball in the bag. But as you can only take off Red Balls 2 by 2 (Did
I already say that ?!) you'll remove the last Blue Balls, one by one......

So the lastball will be red...

Oh, by the way, did I tell you that you take off Red Balls 2 by 2 ?! ;->

For tired people here is why you take off Red Balls 2 by 2 :
- If you take off 1 RED and 1 BLUE, in fact you will take off 1 BLUE
- If you take off 2 RED, in fact you will take off 2 RED (and add 1 BLUE)
- If you take off 2 BLUE, in fact you will take off 1 BLUE

linked list aha:!!

how does one find a loop in a singly linked list in O(n) time using constant
memory? you cannot modify the list in any way (and constant memory
means the amount of memory required for the solution cannot be a function
of n.)

solution: linked list

I first figured this out when I was asked the question in a Microsoft interview,
so there's verification that one question in the book was from a real
interview. The best part of this problem is that I actually needed to write the
code out for it a little while ago to detect a bug.
One way to detect a loop is to iterate over the list with 2 pointers at the
same time where one is iterating at double speed. If the 2 pointers are ever
equal after they iterate once and before they both reach an end, there's a
loop.

Now for another puzzle.. I think the next question I had to answer was
something along the lines of "OK, How do you remove a loop in a linked list
with the same constraints?" The latter question definitely seems harder, but
in the sequence of questions I'd already answered for the interviewer, the
solution was pretty obvious. I'll leave that solution to someone else today.

Hm. I think I should pick up that book. I've always wondered where people
got their juicy interview riddles.. I've always just repeated ones I've heard
before or made up new ones along the same lines.

pennies aha:!!

i challenge you to a game. we each get one penny and we flip them at the
same time. (so on turn 1, we each flip our respective pennies - turn 2, we flip
them again, and so on until someone wins). i am looking to get heads then
tails. you are looking to get heads then heads. so if you flip heads on any flip
and then heads on the next flip, you win. if i flip heads on any flip and then
tails on the next flip, i win. (its not a speed race, we both flip at the same
time, except i'm only concerned with what appears on my coin, and you are
only concerned with whats on your coin). are the odds fair? (obviously not,
otherwise this wouldn't be a question). who has the advantage and why?

solution: pennies

(ummm,, I don't right too good. ;) )

In a 3 turn game:
O needs HH, X needs HT

HHH O

HHT OX

HTH X

HTT X

THH O

THT X

TTH

TTT

X wins 4 times out of 8, O wins 3 times out of 8

Its because in the event X loses there is a 50% chance of starting out on
another H where O has to start on a T 75% of the time. (or something like
that...)

flipping coins aha:!

someone walks into your room and dumps a huge bag of quarters all over
the floor. they spread them out so no quarters are on top of any other
quarters. a robot then comes into the room and is programmed such that if it
sees a head, it flips it to tails. if it sees a tail, it throws it in the air. the robot
moves around randomly forever. will there be a convergence in distribution
of heads vs. tails?

solution: flipping coins

Hmmm. I made a little math trick out of it. If the 'bot finds a head, it flips. If
the bot finds a tail, there's a fifty percent chance this will become a head, as
well.

(P_h = #heads/#coins, P_t = #tails/#coins)

So, delta h = -P_h + .5 P_t

= -(1 - P_t) + .5 P_t

= 1.5 P_t -1

Which is zero when P_t is 2/3. It's important to remember that a flip to a tail
results in no change to the number of tails -- this threw me off for a second.

heaven aha:!

a person dies, and arrives at the gate to heaven. there are three doors. one
of them leads to heaven. another one leads to a 1-day stay at hell, and then
back to the gate, and the other leads to a 2-day stay at hell, and then back
to the gate. every time the person is back at the gate, the three doors are
reshuffled. how long will it take the person to reach heaven?

this is a probability question - i.e. it is solvable and has nothing to do

with religion, being sneaky, or how au dente the pasta might be ;-)

solution: heaven

1/3 of the time, the door to heaven will be chosen, so 1/3 of the time it will
take zero days. 1/3 of the time, the 1-day door is chosen; of those, the right
door will be chosen the next day, so 1/9 trips take 1 day. Similarly, 1/9 will
take two days (choosing the 2-day door, then the right door).

After that, the cases split again, and again, and again. I can't seem to make
a nice infinite sum this way, so let's try again.
Suppose the average days spent is X. 1/3 of the cases are done in zero days
as before. 1/3 of the cases are 1 day plus X. 1/3 are 2 + X. So:

X = 1/3 * 0 + 1/3 * (1 + X) + 1/3 * (2 + X)

= 0 + 1/3 + X/3 + 2/3 + X/3

= 1 + 2X/3

Therefore,

X/3 = 1

X = 3

On average, it takes three days to get to heaven. Two if the noodles are
limp.

Took me one blind alley, and about five minutes. (heh heh)

noodles aha:!

there is a pot of N noodles. (so there are 2N ends). a person randomly grabs
two ends and merges them. the person keeps doing it, until there are no
more noodles, (and only loops), left in the pot. what's the average number of
loops in the pot?

solution: noodles

OK, all the answers so far just list formulae, without giving any explanation.
I'll try to work it out out loud. (At this point I have no idea what the answer
is.) Also, it's a math problem, so I'll assume we can ignore factors like how
much the noodles stick together, how stiff they are, and so on. I'm sure I
have no idea to take account of those factors.

Call the first end he picks up Noodle i. The second end he picks up is Noodle
i*.
When he sticks the first two ends together, there are two possible outcomes:

(a) i* is the other end of noodle i, so he has created a loop.

(ii) i* is a different noodle from i, so he has created one long noodle out of
two.

What are the odds of (a) happening? There are (2n-1) ends left in the bowl
once he picks up the end of noodle i, and only 1 of them is the other end of
the same noodle. Abstracting away from all physical details, let's say the
odds of getting the other end of the same noodle are 1/(2n-1).

So the odds of (b) happening are 1-[1/(2n-1)], which is [(2n-1)-1]/(2n-1),

i.e. (2n-2)/(2n-1).

If (a) happened, now we have a bowl with one loop in it, and n-1 other
unlooped noodles. We add 1 to our count of loops, and repeat the problem
for n-1.

If (b) happened, now we have a bowl with 0 loops in it, and n-1 other
unlooped noodles. It's just that one of those noodles is especially long. We
don't add anything to our count of loops; we just repeat the problem for n-1.

Now, when we get down to 1 unlooped noodle left in the bowl, the odds of
(a) happening are 1.

The average # of loops will be: for each point where a loop could be formed,
add 1 * the probability of a loop being formed then.

So we can write a function:

real average_number_of_loops (int n) {

if (n == 1) {
return 1
} else {
 -- there is a 1/(2n-1) chance of getting 1 loop formed here
-- and a (2n-2)/(2n-1) chance of getting 0 loops formed here
-- and in either case we then have the same problem repeated for
-- n-1 noodles
-- so we should return ([(1/(2n-1))*1] + [(2/(2n-1))*0]) +
average_number_of_loops(n-1)
-- or, simplifying...
return (1/(2n-1)) + average_number_of_loops(n-1)
}

Equivalently:

real average_number_of_loops (int n) {

if (n == 0) {
return 0
} else {
return (1/(2n-1)) + average_number_of_loops(n-1)
}

So I guess I agree with the guy who wrote:

> Summation for i = 1 to N of 1 / 2i - 1. Sorry I can't figure out how to >

resolve.

Except that you have to understand his "1 / 2i - 1" as being "1/ (2i - 1)."
Which is no doubt what he intended. But now we have an explanation of
why.

coin on a table aha:!!!

you die and the devil says he'll let you go to heaven if you beat him in a
game. the devil sits you down at a round table. he gives himself and you a
huge pile of quarters. he says "ok, we'll take turns putting quarters down, no
overlapping allowed, and the quarters must rest on the table surface. the
first guy who can't put a quarter down loses." the devil says he wants to go
first.

being the smart programmer you are, you realize that if the devil goes first,
he may automatically win. so you convince him to let you go first, which
makes your day because you know you can't lose. what is your winning
strategy?

solution: coin on a table

Here's what I came up with:

First, put the first quarter exactly in the center of the (perfectly circular)
table.

Next, for each quarter the opponent places, place one directly opposite it.
That is, place it so that the center of the table is halfway between your piece
and the opponent's previous piece.

This will generate a completely symettric (about the center) layout of

quarters on the table. This means that whenever the opponent selects a free
space to place a quarter in, the space opposite is guaranteed to be free as
well. Since we are always guaranteed an open space, we will never lose with
this strategy (and thus win when there are finally no more spaces for the
opponent to use).

more hat puzzles aha:!

i buried four fishermen up to their necks in the sand on the beach at low tide
for keeping their fishing spot a secret from me. i put a hat on each of their
heads and told them that one of them must shout out the correct color of
their own hat or they will all be drowned by the incoming tide. i give them 10
minutes to do this. fisherman A and B can only see the sand dune i erected.
fisherman C can see that fisherman B has a white hat on. fisherman D can
see that C has a black hat, and B has a white hat. the fisherman have been
told that there are four hats, two white and two black, so they know that
they must have either a white or a black hat on. who shouts out the color of
their hat and how do they know?

solution: more hat puzzles

Fisherman C shouts out.

Fishermen A and B are in the same situation - they have no information to

help them determine their hat colour so they can't answer. C and D realise
this.

Fisherman D can see both B and C's hats. If B and C had the same colour hat
then this would let D know that he must have the other colour.

When the time is nearly up, or maybe before, C realises that D isn't going to
answer because he can't. C realises that his hat must be different to B's
otherwise D would have answered. C therefore concludes that he has a black
hat because he can see B's white one.

cube aha:!

this is difficult to describe in words, so read this carefully, lest there be any
confusion. you have a normal six sided cube. i give you six different colors
that you can paint each side of the cube with (one color to each side). how
many different cubes can you make?
different means that the cubes can not be rotated so that they look the
same. this is important! if you give me two cubes and i can rotate them so
that they appear identical in color, they are the same cube.

solution: cube

Let X be the number of "different" cubes (using the same definition as in the
problem). Let Y be the number of ways you can "align" a given cube in space
such that one face is pointed north, one is east, one is south, one is west,
one is up, and one is down. (We're on the equator.) Then the total number of
possibilities is X * Y. Each of these possibilities "looks" different, because if
you could take a cube painted one way, and align it a certain way to make it
look the same as a differently painted cube aligned a certain way, then those
would not really be different cubes. Also note that if you start with an aligned
cube and paint it however you want, you will always arrive at one of those X
* Y possibilities.

How many ways can you paint a cube that is already "aligned" (as defined
above)? You have six options for the north side, five options for the east
side, etc. So the total number is 6! (that's six factorial, or 6 * 5 * 4 * 3 * 2 *
1). Note that each way you do it makes the cube "look" different (in the
same way the word is used above). So 6! = X * Y.

How many ways can you align a given cube? Choose one face, and point it
north; you have six options here. Now choose one to point east. There are
only four sides that can point east, because the side opposite the one you
chose to point north is already pointing south. There are no further options
for alignment, so the total number of ways you can align the cube is 6 * 4.

Remember, Y is defined as the number of ways you can align the cube, so Y
= 6 * 4. This gives us 6! = X * 6 * 4, so X = 5 * 3 * 2 * 1 = 30.

one mile south aha:!!

how many places are there on the earth that one could walk one mile south,
then one mile east, then one mile north and end up in the same spot? to be
precise, let's assume the earth is a solid smooth sphere, so oceans and
mountains and other such things do not exist. you can start at any point on
the sphere and walk in any direction you like.

think you've figured it out? i'll tell you now, there is more than one. in fact,
there are more than two. also be advised that walking north from the north
pole (or south from the south pole) is illogical and therefore does not enter
into the problem. all normal assumptions about directions will be used.

there are no tricks involved with this question. it just forces you to really
think about the problem to come up with all the solutions.

solution: one mile south aha:

problem: how many places are there on the earth that one could walk one
mile south, then one mile east, then one mile north and end up in the same
spot? to be precise, let's assume the earth is a solid smooth sphere, so
oceans and mountains and other such things do not exist. you can start at
any point on the sphere and walk in any direction you like. think you've
figured it out? i'll tell you now, there is more than one. in fact, there are
more than two. also be advised that walking north from the north pole (or
south from the south pole) is illogical and therefore does not enter into the
problem. all normal assumptions about directions will be used.

there are no tricks involved with this question. it just forces you to really
think about the problem to come up with all the solutions.

solution:

well the north pole is one such place.

then somewhere near the south pole such that when you walk one mile south
you are at the point on the earth where the circumference is 1. that way
when you walk 1 mile east, you end up back at the same point. and of course
one mile north from there puts you back where you started. here is a
drawing courtesy of jy. there may or may not be such a place in the northern
hemisphere where walking a mile south puts you at the 1 mile circumference
point on the earth.

i'm no geometry sphere expert, so someone will have to let me know if that
is physically possible (i.e. i tend to think that if you walk n units south from
any point on the northern part of a sphere, other than the north pole, it is
impossible for the circumference to be n or less than n, but who knows?)

finally there are actually an infinite number of points. if we consider the case
before where we went to the point with a circumference of 1, why not go to
the point with a circumference of 1/2. then when you go a mile east, you
loop around twice, and end up in the same spot. this holds true for 1/3, 1/4,
1/5, ... 1/n, etc.
ants on a triangle aha:!

there are three ants on a triangle, one at each corner. at a given moment in
time, they all set off for a different corner at random. what is the probability
that they don't collide?

solution: ants on a triangle

Consider the triangle ABC. We assume that the ants move towards different
corners along the edges of the triangle.

Total no. of movements: 8 A->B, B->C, C->A A->B, B->A, C->A A->B, B-
>A, C->B A->B, B->C, C->B A->C, B->C, C->A A->C, B->A, C->A A->C, B-
>A, C->B A->C, B->C, C->B

Non-colliding movements: 2 A->B, B->C, C->A A->C, B->A, C->B

(i.e. the all ants move either in the clockwise or anti-clockwise direction at
the same time)

P(not colliding) = 2/8 = 0.25

screwy pirates aha:!

a. first i would just like to say i am so glad jenna got voted off survivor. yuck,
i couldn't stand her ever since she accused that guy of eating beef jerky
when he was chewing on grass.
b. recruit meyer on boot camp rocks. did you see how he made himself cry to
get sympathy? and it worked! he is the funniest part of the show. i hope he
wins.
c. i apologize for this site being soooooo slow. but that's why you get what
you pay for. (i didn't pay anything so i shouldn't really expect anything in
return.)
d. those screwy pirates are at it again. this time there are 13 pirates and
they need to protect their treasure chest. they decide that they should only
be able to open the chest if the majority (at least 7) agree that it should be
opened. they ask a locksmith to come and put a specific number of locks on
the safe. every lock must be opened to open the chest. there can be multiple
keys for each lock, but each key only opens one lock (i.e. no skeleton keys).
the locksmith can give more than one key to each pirate. how many locks
should the locksmith use and what strategy should he use to distribute the
keys, such that only when a majority of the pirates agree can the chest be
opened?

solution: screwy pirates aha:

problem: those screwy pirates are at it again. this time there are 13 pirates
and they need to protect their treasure chest. they decide that they should
only be able to open the chest if the majority (at least 7) agree that it should
be opened. they ask a locksmith to come and put a specific number of locks
on the safe. every lock must be opened to open the chest. there can be
multiple keys for each lock, but each key only opens one lock (i.e. no
skeleton keys). the locksmith can give more than one key to each pirate.
how many locks should the locksmith use and what strategy should he use to
distribute the keys, such that only when a majority of the pirates agree can
the chest be opened?

a somewhat technical answer is here.

(un)fortunately that user left it up to me to explain why this solution works...

but fortunately my brother did it here in a very understandable fashion.
thanks big brother!

technical solution for screwy pirates

Here is solution that works for any number of pirates and any number of
pirates needed to open to chest.
Let T be the total number of pirates. Let N be the number of pirates required
to open the chest.

The number of locks needed would be L = (T,N-1) = (T!)/[(N-1)! * (T-N-1)!].

The number of keys each pirate would have be K = (T-1,N-1) = (T-1)!/[(N-
1)! * (T-N)!].

For this specific problem, the number of locks would be 1716 and the number
of keys per pirate would be 924.

I know I haven't provided an explanation of why and how this system works.
We'll leave that lengthy, involved explanation to the author.

Note: the notation (x,y) is the combination notation; I can't use the
conventional combination notation in plain text. (x,y) basically asks the
question "how many ways can you pick y objects from a group of x objects?"
(x,y) = (x!)/[y! * (x-y)!]

easy solution for screwy pirates

Here's a less elegant solution. How many different combinations of 6 pirates

are there? 13 choose 6 is: 13!/6!7! That's a hell of a lot of locks. But you
could say, put that many locks on the chest. And for each lock, you distribute
keys to all the pirates EXCEPT the selection of 6 corresponding to that lock.
Then any selection of 7 pirates is guaranteed to be open all the locks. For
consider the first 6 of them. There will be exactly ONE lock that those six are
unable, between the 6 of them, to open. But the seventh pirate will be able
to open that lock. Because keys to that lock went out to everybody except
those 6.

Also, any selection of 6 pirates is guaranteed to encounter exactly one lock

that they are unable to open. (So any selection of fewer than 6 is guaranteed
to encounter 1 or more locks they are unable to open...)
So that will do the trick, and the explanation of why it would do the trick is
elegant. But I don't feel that it's an elegant solution, because of there being
so many locks... Anyone have something better?

treasure island aha:!

you find an old treasure map in your grandma's attic. the map shows a
cannon, a coconut tree, and a palm tree. the map states that to find the
treasure you must:
a. start at the cannon, walk toward the palm tree while counting your paces.
when you reach the palm tree, turn 90 degrees to your left and walk the
same number of paces. mark that spot on the ground with a stake.
b. start at the cannon again, walk toward the coconut tree while counting
your steps. when you reach the coconut tree, turn 90 degrees to your right
and walk the same number of paces. mark that spot on the ground with a
stake.
c. find the midpoint between the two stakes and dig for the treasure.

you set off in secrecy to the deserted island. upon reaching the shore you
site the coconut tree and the palm tree, but someone has removed the
cannon. without digging randomly all over the island, is it still possible to find
the treasure?

solution: treasure island

problem: you find an old treasure map in your grandma's attic. the map
shows a cannon, a coconut tree, and a palm tree. the map states that to find
the treasure you must:
a. start at the cannon, walk toward the palm tree while counting your paces.
when you reach the palm tree, turn 90 degrees to your left and walk the
same number of paces. mark that spot on the ground with a stake.
b. start at the cannon again, walk toward the coconut tree while counting
your steps. when you reach the coconut tree, turn 90 degrees to your right
and walk the same number of paces. mark that spot on the ground with a
stake.
c. find the midpoint between the two stakes and dig for the treasure.

you set off in secrecy to the deserted island. upon reaching the shore you
site the coconut tree and the palm tree, but someone has removed the
cannon. without digging randomly all over the island, is it still possible to find
the treasure?

solution: this just takes basic geometry skills. when we get to the island all
we see are the coconut and palm trees. so lets lay out our coordinate system
such that the palm tree is at (0,0) and the coconut tree is at (1,0). it
honestly doesn't matter how you describe the coordinate system - you could
say the coconut is at (c,0) if you like, or even (0,c) or (0,1). we are just
placing our own coordinate system on top of the existing surface. if you use a
different system, you will get a different numerical answer but the same
positional answer in the real world.

here is our island. the cannon is at (x,y) because we have no idea where it is,
so x and y are the unknowns. (note the cannon doesn't have to be in the
upper right quadrant but it won't make a difference in the solution because x
and y could be negative if we want them to be).
if we walk to the palm and turn right we can easily see the way it lays out in
the diagram. you basically have just transposed the x and y positions.

it might be easier to think of it as a triangle, and flipping the triangle around

to find the point for the next stake at (y, -x)
we do the same thing with the coconut tree, although here its only a tiny bit
trickier because we have to factor in the position of the coconut tree at (1,0)
or (c,0).

we use the idea of a triangle to help us better understand that the stake will
end up at (1-y, x+1) or (c-y, x+c).

then to find the midpoint of the two points which is just the first position +
the second position divided by two.

(y, -x) and (1-y, x+1)

((y - y + 1)/2, (x - x + 1)/2)

(1/2, 1/2)
hence our answer is (1/2,1/2) - although we'll see that if we had use the
constant C for the coconut tree we would have ended up with (c/2, c/2). this
is important because even though we laid out the island in our own
coordinate system, its not always the case that c must be positive. i think
(and i've seen this solution elsewhere, but can't really come up with a repro,
so let me know if i'm wrong here) there really are two places the treasure
could be because c could also be negative. so if we use our first answer of
(1/2,1/2) we must also search at (-1/2, -1/2). if the cannon was in the lower
half of the coordinate system you can see that the treasure would actually be
in that quadrant also.

mountain man aha:!

at 6 a.m. a man starts hiking a path up a mountain. he walks at a variable

pace, resting occasionally, but never actually reversing his direction. at 6
p.m. he reaches the top. he camps out overnight. the next morning he wakes
up at 6 a.m. and starts his descent down the mountain. again he walks down
the path at a variable pace, resting occassionally, but always going downhill.
at 6 p.m. he reaches the bottom. what is the probability that at some time
during the second day, he is in the exact same spot he was in on the first
day?

solution: mountain man

problem: at 6 a.m. a man starts hiking a path up a mountain. he walks at a

variable pace, resting occasionally, but never actually reversing his direction.
at 6 p.m. he reaches the top. he camps out overnight. the next morning he
wakes up at 6 a.m. and starts his descent down the mountain. again he
walks down the path at a variable pace, resting occassionally, but always
going downhill. at 6 p.m. he reaches the bottom. what is the probability that
at some time during the second day, he is in the exact same spot he was in
on the first day?
answer: the probability is 100%. the easiest way to see it is, consider that on
the second day when the man is going down the mountain, a ghost follows
his original pace up the mountain. so even if he varies his pace as he goes
down the mountain, at some point in time, he will be in the same spot as the
ghost, and therefore, the same spot he was in the day before.

clock aha:!

part I: what is the angle between the minute hand and the hour hand at 3:15
on an analog clock? no, its not 0.

part II: how often does the minute hand pass the hour hand on an analog
clock?

solution: clock

part I: what is the angle between the minute hand and the hour hand at 3:15
on an analog clock? no, its not 0.

part II: how often does the minute hand pass the hour hand on an analog
clock?

answer: part I: 12 hours on the clock make 360 deg. so one hour is 30 deg.
the hour hand will be directly on the 3 when the minute hand is at 12 (3:00).
after 15 minutes or 1/4 of an hour, the hour hand will be 1/4 * 30 deg = 7.5
deg. away from the minute hand.

part II: if you just think about it, intuitively you'll see the minute hand
passes the hour hand 11 times every 12 hours, so it must pass it every 1
1/11 hours. but this doesn't make sense to me. i need to prove it.

if x is our answer then every x hours, the minute hand and the hour hand will
be right on top of each other. every hour the hour hand travels 5 units. so
between every time that the minute and the hour hand meet, the hour hand
will go 5*x units. every hour the minute hand travels 60 units, so it will have
gone 60*x units.

what we're trying to find is the distance traveled by the minute hand to reach
the hour hand, once the minute hand has looped around once. consider its
12:00. both hands in the same position. after an hour, minute hand is on 12,
hour hand on 1 (its traveled 5 units). now in the time it takes the minute
hand to catch up to the hour hand it will travel a little bit further.

we only need to find x where 5*x = 60*(x-1), since the real distance
traveled by the minute hand, from where it started to where it ends, is
60*(x-1). the first hour just puts it back where it started, so we're only
concerned with the extra part it traveled to reach the hour hand.

5x = 60(x-1)

5x = 60x - 60

60 = 55x

60/55 = x

there it is. the answer is 60/55 hours, or every 1 and 1/11 hours.

i apologize that this is horribly confusing, but if you stare at it long enough it
will make sense.

surgeons aha:!!

a one armed surgeon with a hand wound needs to operate on three patients.
the surgeon only has two gloves. how can he operate on the three patients in
turn without risking exchange of fluids? (remember he only has one arm so
he only needs to wear one glove at a time.)

solution: surgeons

problem: a one armed surgeon with a hand wound needs to operate on

three patients. the surgeon only has two gloves. how can he operate on
the three patients in turn without risking exchange of fluids? (remember
he only has one arm so he only needs to wear one glove at a time.)

solution: the surgeon places both gloves on his hand (1 and 2). he
operates on patient A. he then takes the top glove off (#2), leaving on the
bottom glove (#1) and operates on patient B. then he carefully reverses
glove #2, so the clean side is on the outside, and he places it on top of
glove #1 which is on his hand, and operates on patient C.

this problem is kind of dumb because how's the surgeon going to change
the gloves on his hand when he only has one hand. plus no offense, but
how often do you come across a one-armed surgeon (i'm sure there are
plenty of one-armed doctors, but a surgeon!?!). anyway, i had to make
this problem child friendly and changing the story to the above was the
only way to do it. consider for a minute what the initial problem was. the
surgeon was just a guy, the patients were women, and the glove was...
well, i won't insult your intelligence.

gold chain aha:!!!

a man has a gold chain with 7 links. he needs the service of a laborer for 7
days at a fee of one gold link per day. however, each day of work needs to
be paid for separately. in other words, the worker must be paid each day
after working and if the laborer is ever overpaid he will quit with the extra
money. also he will never allow himself to be owed a link.

what is the fewest # of cuts to the chain to facilitate this arrangement and
how does that guarantee payment?
solution: gold chain

Can we get change back from the laborer?

If so, we cut one link to make a chain of 4 links, a chain of 2 links and the
cut link itself.

Day 1, we give him the cut link

Day 2, we take back the cut link, give him the 2 link chain
Day 3, we give him the cut link
Day 4, we take back both the cut link and the 2 link chain, give him the 4
link chain
Day 5, we give him the cut link
Day 6, we take back the cut link, give him the 2 link chain
Day 7, we give him the cut link

monty hall problem aha:!

another well known problem in probability is the monty hall problem.

you are presented with three doors (door 1, door 2, door 3). one door has a
million dollars behind it. the other two have goats behind them. you do not
know ahead of time what is behind any of the doors.

monty asks you to choose a door. you pick one of the doors and announce it.
monty then counters by showing you one of the doors with a goat behind it
and asks you if you would like to keep the door you chose, or switch to the
other unknown door.

should you switch? if so, why? what is the probability if you don't switch?
what is the probability if you do.

lots of people have heard this problem.. so just knowing what to do isn't
sufficient. its the explanation that counts!
solution: monty hall problem

another well known problem in probability is the monty hall problem.

you are presented with three doors (door 1, door 2, door 3). one door has a
million dollars behind it. the other two have goats behind them. you do not
know ahead of time what is behind any of the doors.

monty asks you to choose a door. you pick one of the doors and announce it.
monty then counters by showing you one of the doors with a goat behind it
and asks you if you would like to keep the door you chose, or switch to the
other unknown door.

should you switch? if so, why? what is the probability if you don't switch?
what is the probability if you do.

lots of people have heard this problem.. so just knowing what to do isn't
sufficient. its the explanation that counts!

the answer is that yes, you should *always* switch as switching increases
your chances from 1/3 to 2/3. how so, you ask? well, lets just enumerate the
possibilities.

door 1 door 2 door 3

case 1 $$ goat goat

case 2 goat $$ goat

case 3 goat goat $$

its clear that if you just choose a door and stick with that door your chances
are 1/3.

using the switching strategy, let's say you pick door 1. if its case 1, then you
lose. if it's case 2, monty shows you door 3, and you switch to door 2, you
win. if it's case 3, monty shows you door 2, and you switch to door 3, you
win. it doesn't matter what door you pick in the beginning, there are always
still three possibilities. one will cause you to lose, and two will cause you to
win. so your chances of winning are 2/3.

the solution all resides in the fact that monty knows what is behind all the
doors and therefore always eliminates a door for you, thereby increasing
your odds.

maybe its easier to see in this problem. there are 1000 doors, only one of
which has a prize behind it. you pick a door, then monty opens 998 doors
with goats behind them. do you switch? it seems more obvious in this case,
because monty had to take care in which door not to open, and in the
process basically showing you where the prize was (999 out of 1000 times).

classic weighing aha:!

this is a classic problem which i have heard many times before. this is the
"harder" of the two problems, since in this one, you do not know if the invalid
item weighs more or less than the others.

solving it is only half the battle. writing up a solution that anyone including
your grandma could understand, is very hard.

problem: the evil king from before sends his own assassin to take care of the
evil queen who tried to poison him. of course, her trusty guards catch the
assassin before any harm is done. the queen notices that the assassin is
quite handsome and doesn't really want to punish him by death. she decides
to test his wisdom.

the queen gives the assassin 12 pills which are all completely identical in
shape, smell, texture, size, except 1 pill has a different weight. the queen
gives the man a balance and tells him that all the pills are deadly poison
except for the pill of a different weight. the assassin can make three
weighings and then must swallow the pill of his choice. if he lives, he will be
sent back to the bad king's kingdom. if he dies, well, thats what you get for
being an assassin.

only one pill is not poison and it is the pill which has a different weight. the
assassin does not know if it weighs more or less than the other pills. how can
he save his skin?

solution: classic weighing

this is a classic problem which i have heard many times before. this is the
"harder" of the two problems, since in this one, you do not know if the invalid
item weighs more or less than the others.

solving it is only half the battle. writing up a solution that anyone including
your grandma could understand, is very hard.

problem: the evil king from before sends his own assassin to take care of the
evil queen who tried to poison him. of course, her trusty guards catch the
assassin before any harm is done. the queen notices that the assassin is
quite handsome and doesn't really want to punish him by death. she decides
to test his wisdom.

the queen gives the assassin 12 pills which are all completely identical in
shape, smell, texture, size, except 1 pill has a different weight. the queen
gives the man a balance and tells him that all the pills are deadly poison
except for the pill of a different weight. the assassin can make three
weighings and then must swallow the pill of his choice. if he lives, he will be
sent back to the bad king's kingdom. if he dies, well, thats what you get for
being an assassin.

only one pill is not poison and it is the pill which has a different weight. the
assassin does not know if it weighs more or less than the other pills. how can
he save his skin?
solution: easy.

choose any eight of the pills and put four of them on each side of the
balance.

there are two possibilities:

(1) one side of the balance comes out lighter. In this case, you know that the
abnormal (safe) pill is one of the pills already on the balance. label the pills
on the lighter side A B C and D, and the pills on the heavier side E F G and H.
label the pills not on the balance NORM (you know they’re normal pills).

(2) the balance is even. in this case, you know that the abnormal (safe) pill is
one of the pills not on the balance. label the pills already on the balance
NORM, and label the four pills not on the balance I J K and L.

let’s proceed with possibility (1).

consider why the side ABCD came out higher than the side EFGH. this could
be because:

A is the abnormal pill, and it’s lighter than the other pills.

B is the abnormal pill, and it’s lighter than the other pills.

C is the abnormal pill, and it’s lighter than the other pills.

D is the abnormal pill, and it’s lighter than the other pills.

E is the abnormal pill, and it’s heavier than the other pills.

F is the abnormal pill, and it’s heavier than the other pills.

G is the abnormal pill, and it’s heavier than the other pills.

H is the abnormal pill, and it’s heavier than the other pills.
now let’s make another weighing, with two of the ABCD pills on either side,
and one of the EFGH pills on either side. for example, let’s weigh ABE versus
CDF. how would this weighing come out given each of those 8 possibilities we
just listed?

if A is the light pill, the ABE/CDF weighing will come out with ABE high.

if B is the light pill, the ABE/CDF weighing will come out with ABE high.

if C is the light pill, the ABE/CDF weighing will come out with ABE low.

if D is the light pill, the ABE/CDF weighing will come out with ABE low.

if E is the heavy pill, the ABE/CDF weighing will come out with ABE low.

if F is the heavy pill, the ABE/CDF weighing will come out with ABE high.

if G is the heavy pill, the ABE/CDF weighing will come out even.

if H is the heavy pill, the ABE/CDF weighing will come out even.

OK, so we observe how the ABE versus CDF weighing actually comes out.

(a) if it comes out even, then we know that the abnormal pill is either G or H.
for our third weighing, we can weigh G against one of the pills we already
know to be normal (one of the pills we labelled NORM). if it comes out even,
then G is normal and H must be the abnormal pill. if it comes out uneven,
then G is the abnormal pill.

(b) as we can see from our chart above, if the ABE/CDF weighing comes out
with ABE high, then the situation is either: A is the light pill, B is the light pill,
or F is the heavy pill.

(c) as we can see from our chart above, if the ABE/CDF weighing comes out
with ABE low, then the situation is either: C is the light pill, D is the light pill,
or E is heavy pill.
so in either situation (b) or (c), we have two possible light pills and one
possible heavy pill. what we do in that case is we put one of the possible light
pills and the possible heavy pill on one side of the scale, and two NORM pills
on the other side of the scale. this is our third weighing. if it comes out even,
then we know that the other possible light pill is the abnormal pill. if it comes
out with the two NORM pills high, then we know that one of the pills on the
other side is abnormally heavy, so we know that the possible heavy pill is the
culprit. if it comes out with the two NORM pills low, then we know that one of
the pills on the other side is abnormally light, so we know that the possible
light pill on the scale is the culprit.

that takes care of case (1), where the first weighing came out uneven.

what about case (2), where the first weighing comes out even?

then we know the abnormal pill is one of I J K or L, and we have two

weighings to find the abnormal pill in.

for our second weighing, we put I and J on one side of the scale, and two
NORM pills on the other.

(a) if this comes out uneven, we know the abnormal pill is I or J; we weigh I
against one NORM pill to see if I is abnormal and if it isn’t, we can conclude
that J is the abnormal pill.

(b) if the IJ versus 2 NORM weighing comes out even, we know the abnormal
pill is K or L; we weight K against one NORM pill to see if K is abnormal and if
it isn’t, we can conclude that L is the abnormal pill.

finished.

hard river crossing aha:!

a disfunctional family has to cross the river. on one side of the river are a
mom and 2 daughters, dad and 2 sons, the maid and the dog. there is a boat
only big enough to hold 2 people (counting the dog as 1 person). only the
adults are capable of operating the boat. everyone has to get to the other
side, without anything bad happening.

difficulties: if the dog is left with anyone and the maid isn't there to control
him, he'll bite. the dad can't be left with any of the daughters when the mom
isn't there. likewise, the mom can't be trusted alone with either of the sons
when the dad isn't there.

remember! only an adult can operate the boat, AND the boat can't drive
itself.

solution: hard river crossing

a disfunctional family has to cross the river. on one side of the river are a
mom and 2 daughters, dad and 2 sons, the maid and the dog. there is a boat
only big enough to hold 2 people (counting the dog as 1 person). only the
adults are capable of operating the boat. everyone has to get to the other
side, without anything bad happening.

difficulties: if the dog is left with anyone and the maid isn't there to control
him, he'll bite. the dad can't be left with any of the daughters when the mom
isn't there. likewise, the mom can't be trusted alone with either of the sons
when the dad isn't there.

remember! only an adult can operate the boat, AND the boat can't drive
itself.

solution:
we start with a mother (m), two daughters (d1, d2), a father (f), two sons
(s1, s2), a housemaid (h), and a dog (c - canine) on the west (W) shore, and
they all want to get to the east (E) shore.

W = {m, d1, d2, f, s1, s2, h, c} // everyone on the west shore

E = {} // no one on the east shore

let's move everyone, over...

housemaid and canine go east, and the housemaid comes back:

W = {m, d1, d2, f, s1, s2, h}

E = {c}

housemaid and s1 go east, h and c come back:

W = {m, d1, d2, f, s2, h, c}

E = {s1}

father and s2 go east, father comes back:

W = {m, d1, d2, f, h, c}

E = {s1, s2}

mother and father go east, mother comes back:

W = {m, d1, d2, h, c}

E = {f, s1, s2}

h and c go east, father comes back:

W = {m, d1, d2, f}

E = {s1, s2, h, c}

father and mother go east, mother comes back:

W = {m, d1, d2}

E = {f, s1, s2, h, c}

mother and d1 go east, housemaid and c come back:

W = {d2, h, c}

E = {m, d1, f, s1, s2}

h and d2 go east, h comes back

W = {h, c}

E = {m, d1, d2, f, s1, s2}

h and c go east

W = {}

E = {m, d1, d2, f, s1, s2, h, c}

webloggers aha:!

five webloggers - joshua Allen, meg Hourihan, jason Kottke, robert Scoble,
and joel Spolsky - were competing for karma points on the major search
engines: google, yahoo, altavista, lycos, and msn. karma was distributed on
a five point scale. the most popular weblog received 5 points, and the least
popular received 1 point. for each search engine, no two webloggers
received the same number of points. overall scores were determined by
adding up the individual scores from each search engine.

Allen got the highest number of karma points - 24. Kottke was consistent in
his scores: he got the same karma points from 4 different search engines.
Spolsky got 5 points from lycos, and 3 from msn.

no two webloggers got the same total score, and the final rankings were as
follows: Allen, Hourihan, Kottke, Scoble, and Spolsky. how many karma
points did Hourihan get from lycos?

solution: webloggers

five webloggers - joshua Allen, meg Hourihan, jason Kottke, robert Scoble,
and joel Spolsky - were competing for karma points on the major search
engines: google, yahoo, altavista, <>lycos, and msn. karma was distributed
on a five point scale. the most popular weblog received 5 points, and the
least popular received 1 point. for each search engine, no two webloggers
received the same number of points. overall scores were determined by
adding up the individual scores from each search engine.

Allen got the highest number of karma points - 24. Kottke was consistent in
his scores: he got the same karma points from 4 different search engines.
Spolsky got 5 points from lycos, and 3 from msn.

no two webloggers got the same total score, and the final rankings were as
follows: Allen, Hourihan, Kottke, Scoble, and Spolsky. how many karma
points did Hourihan get from lycos?

solution: let's start with what we know

G Y A L M Total
|============================
A | 24
H |
K |
Sc|
Sp| 5 3

the only possible values for Allen achieving 24 is { 5 5 5 5 4 } and since

Spolsky got a 5 from lycos, we know that is where Allen's 4 comes from.

we also know that the total number of points given out was 75.
(5 * (5 + 4 + 3 + 2 + 1))

spolsky had to have at least 11 points. if Spolsky had more than 11 points,
say 12, then is it possible to achieve a solution? Scoble would have had to
have at least 13 (since there were no ties), and Kottke 14, and Houlihan 15.
that would yield an overall total of 78. too much! so Spolsky definitely had 11
points.

G Y A L M Total
|============================
 A | 5 5 5 4 5 24
H |
K |
Sc|
Sp| 1 1 1 5 3 11

using the same logic as before, we also know that Scoble could not have
gotten more than 12 points. if he had 13, and Kottke 14, and Houlihan 15,
the total would be 77. still too much. so Scoble had 12, and continuing on
Kottke had to have 13 and Houlihan 15, otherwise the totals would be over
75.

now we know Kottke had 14 points. if he got four 4's for consistency, it
wouldn't work (already over 16). if he got four 2's, it also wouldn't work (8
points plus the maximum 5 is still only 13). so he had to have received four
3's. and since he couldn't have gotten a 3 from msn, that is where he
received a 1.

G Y A L M Total
|============================
A | 5 5 5 4 5 24
H | 15
K | 3 3 3 3 1 13
Sc| 12
Sp| 1 1 1 5 3 11

let's look at scoble. we can see from the chart that all 5's and 3's have
already been given out (and there is only one 1 left). so Scoble's scores can
only contain 4's, 2's, or a single 1. given that information the only possible
combination of 5 scores that would yield 12 is { 2 2 2 2 4 }. since Allen
already has a 4 from lycos, Scoble must have a 2 there.

G Y A L M Total
|============================
A | 5 5 5 4 5 24
H | 15
K | 3 3 3 3 1 13
Sc| 2 12
 Sp| 1 1 1 5 3 11

hence Houlihan must have a 1 from lycos!

G Y A L M Total
|============================
A | 5 5 5 4 5 24
H | 1 15
K | 3 3 3 3 1 13
Sc| 2 12
Sp| 1 1 1 5 3 11

shapes aha:!!

part I: draw a square. divide it into four identical squares. remove the
bottom left hand square. now divide the resulting shape into four identical
shapes.

part II: draw an equilateral triangle (all sides same length). divide it into four
identical shapes. remove the bottom left hand shape. now divide the
resulting shape into four identical shapes.

this is the sort of problem that i would expect on a MENSA test. i'm not too
sure whether getting this right constitutes intelligence in a way that would
benefit computer scientists, but maybe it does. if you figure it out, then you
can say it does. if you can't figure it out, then you can just say it's all
hogwash and it's a stupid question.

solution: shapes

part I: draw a square. divide it into four identical squares. remove the
bottom left hand square. now divide the resulting shape into four identical
shapes.

part II: draw an equilateral triangle (all sides same length). divide it into four
identical shapes. remove the bottom left hand shape. now divide the
resulting shape into four identical shapes.
this is the sort of problem that i would expect on a MENSA test. i'm not too
sure whether getting this right constitutes intelligence in a way that would
benefit computer scientists, but maybe it does. if you figure it out, then you
can say it does. if you can't figure it out, then you can just say it's all
hogwash and it's a stupid question.

solution:

part I

part II
easy river crossing aha:!

three cannibals and three anthropologists have to cross a river. the boat they
have is only big enough for two people. if at any point in time there are more
cannibals on one side of the river than anthropologists, the cannibals will eat
them. what plan can the anthropologists use for crossing the river so they
don't get eaten?

remember! the boat can't cross the river by itself, someone has to be in it to
row it across.

a much harder river crossing problem will appear later this week.

A - anthropologist

C - cannibal

++ - boat

river

AAA |============|

|++ |

CCC |============|

need to make it

river

|============| AAA

| ++|

|============| CCC

note that if you violate the "anthropologists > cannibals" rule at any point in
time, it is illegal.. for example if a boat with a cannibal and an anthropologist
travels to a shore with one cannibal on it, then # cannibals > #
anthropologists, even if you say the anthropologist immediately takes the
boat back.
solution: easy river crossing

problem: three cannibals and three anthropologists have to cross a river. the
boat they have is only big enough for two people. if at any point in time
there are more cannibals on one side of the river than anthropologists, the
cannibals will eat them. what plan can the anthropologists use for crossing
the river so they don't get eaten?

A - anthropologist

C - cannibal

++ - boat

river

AAA |============|

|++ |

CCC |============|

need to make it

river

|============| AAA

| ++|

|============| CCC

note that if you violate the "anthropologists > cannibals" rule at any point in
time, it is illegal.. for example if a boat with a cannibal and an anthropologist
travels to a shore with one cannibal on it, then # cannibals > #
anthropologists, even if you say the anthropologist immediately takes the
boat back.

Let W be the west shore which they are all on. Let E be the east shore where
they want to go.

1. A and C cross
W = { A, A, C, C }

E = { A, C }

2. A returns

W = { A, A, A, C, C }

E = { C }

3. C and C cross

W = { A, A, A }

E = { C, C, C }

4. C returns

W = { A, A, A, C }

E = { C, C }

5. A and A cross

W = { A, C }

E = { A, A, C, C }

6. A and C return

W = { A, A, C, C }

E = { A, C }

7. A and A cross

W = { C, C }

E = { A, A, A, C }

8. C returns

W = { C, C, C }

E = { A, A, A }

9. C and C cross
W = { C }

E = { A, A, A, C, C }

10. C returns

W = { C, C }

E = { A, A, A, C }

11. C and C cross

W = true

E = { A, A, A, C, C, C }

chessboard aha:!!!

problem: using 31 dominoes, where one domino covers exactly two squares,
can you cover all the empty squares on this chessboard (which has 62
spaces). if so, how? if not, why?

solution: chessboard

problem: using 31 dominoes, where one domino covers exactly two

squares, can you cover all the empty squares on this chessboard (which
has 62 spaces). if so, how? if not, why?
solution: i think everyone's first inclination is to try and figure out how it is
possible. then again, if you've heard a bunch of these questions before,
you usually know that if the question says "if not, why?" or "prove
whether its possible or impossible", you can infer that it is not possible
(otherwise, the question usually just asks for the solution).

after awhile fiddling around with the dominoes, you will start to realize
that the most obvious solutions are not viable. if you start to think its
impossible, think about why.

a reader emailed me telling me that using the picture of the chessboard

without the squares colored will result in a very small percentage of people
solving the problem. if i use the following picture

a much higher percentage of people will be able to solve the problem. the
people who solve the problem using only the grid, usually represent the
problem in their heads as the colored board.

still can't figure it out? using the colored board, remember that a domino
always covers a black square and a white square.

if you look at the board, you will see that the two squares missing are
both black. this means that for all the squares on the board, each white
square has a corresponding black square, except for two white ones. so
even if you covered all the black/white pairs, you would still be left with
two white squares, which will never be adjacent to each other, no matter
how you lay out the dominoes.

dave's on fire aha:

dave winer is stuck on a deserted island, with lots of trees, which is very thin
and ten miles long (east to west). large cliffs surround the entire island and if
he jumped off, he wouldn't survive the fall. a fire starts burning at the west
side of the island. unfortunately this island always has a west to east blowing
wind blowing at 2mph and this moves the fire slowly toward dave at 1mph.
(so he only has ten hours left). save dave (or maybe, let him burn :-) ! what
to do?

solution: dave's on fire

someone suggested he could dig a pit across the island to act as a firebreak.
good suggestion, if he had a shovel and the ground wasn't too hard.

but even if he didn't have a shovel, he could pick up a branch and run up to
the fire and light the branch. then run all the way to the eastern edge of the
island, but stop about a mile short. there he could light all those trees on fire
and they would start burning and the fire would move east. it would consume
all that vegetation in an hour, and then dave could wait for awhile for that
part to cool down some. when the initial fire reached him, he could just run
into the already burnt part, and the fire couldn't get him.

T = tree

D = dave

B = burnt

top view

==========================

| fire-> T T T T T T T D |

==========================

dave gets branch and lights it from fire

==========================

| fire-> D T T T T T T T |

==========================

dave lights trees farther east on the island

==========================

| fire-> T T T T D fire->|

==========================

dave waits until second fire cools, and then hides out there

==========================

| B B B fire-> T T B B D |

==========================

dave is saved!
fuse on fire aha:!!

a mad bomber is out on the job, making bombs. he has two fuses (pieces of
string) of varying thickness which each burn for 30 seconds. unfortunately he
wants this bomb to go off in 45 seconds. he can't cut the one fuse in half
because the fuses are different thicknesses and he can't be sure how long it
will burn. how can he arrange the fuses to make his bomb go off at the right
time?

solution: fuse on fire

problem: a mad bomber is out on the job, making bombs. he has two fuses
(pieces of string) of varying thickness which each burn for 30 seconds.
unfortunately he wants this bomb to go off in 45 seconds. he can't cut the
one fuse in half because the fuses are different thicknesses and he can't be
sure how long it will burn. (for example: the first half of the fuse might burn
up in 10 seconds and the second half in 20 seconds.. the fuse doesn't burn at
a constant rate, but the total time it would burn is 30 seconds). how can he
arrange the fuses to make his bomb go off at the right time?

solution: light both ends of one of the fuses. when that fuse goes out, 15
seconds has elapsed. then light the other fuse.

pirates revisited aha:!

a slightly different version of the original pirates problem (read that one first
to get all the rules). 6 pirates, only one gold coin. as before, the pirates are
super-smart, and they value, in this order: (i) their lives, (ii) getting money,
(iii) seeing other pirates die. so if given the choice between two outcomes, in
which they get the same amount of money, they'd choose the outcome
where they get to see more of the other pirates die. how can pirate 6 save
his skin?

solution: pirates revisited

problem: a slightly different version of the original pirates problem (read that
one first to get all the rules). 6 pirates, only one gold coin. as before, the
pirates are super-smart, and they value, in this order: (i) their lives, (ii)
getting money, (iii) seeing other pirates die. so if given the choice between
two outcomes, in which they get the same amount of money, they'd choose
the outcome where they get to see more of the other pirates die. how can
pirate 6 save his skin?1 pirate case is obvious. 2 pirate case is obvious,
pirate 2 keeps the coin.

3 pirate case: pirate 3 has to give the coin to pirate 1, because if he gives it
pirate 2 pirate 2 will say "screw that i wanna see you die and i'm going to get
the coin anyway." so in 3 pirate case, we have pirate 1 gets 1 coin, pirates 2
and 3 get 0.

4 pirate case: pirate 4 can't give the coin to pirate 1, because pirate 1 would
rather see him die since he's going to get 1 coin anyway. But pirate 4 could
give the coin to either pirate 2 or pirate 3.

5 pirate case: pirate 5 just dies, and it goes to the 4 pirate case. there is no
way for him to convince two people to vote for him.

6 pirate case: pirate 6 can count on his own vote, plus pirate 5's vote,
because 5 won't want to die. now who should he give the coin to? he could
give it to pirate 1 or to pirate 4, since in the 4 pirate case they are
guaranteed to get nothing. it's unclear whether he could give it to pirate 2 or
3. neither pirate 2 nor pirate 3 is guaranteed to get a coin in the 4 pirate
case. so the question is, how do they value (i) definitely getting a coin from
pirate 6 vs. (ii) definitely seeing pirates 6 and 5 die, with the chance of
getting a coin from pirate 4. since we don't have enough information to
answer that, to be safe, i would just say pirate 6 should offer the coin to
pirate 1 or pirate 4.

card trick without the trick aha:!!

this is a card trick without the trick. there is no sleight of hand, no tricks up
my sleeve, no magic whatsoever. it's all done with logic, yet it will amaze
most people.

joel and i are working together as a team to do the trick. babak will be the
culprit.

i ask babak to pick 5 cards out of a deck. he can pick any five cards, he can
shuffle the deck 7 times, it really doesn't matter. he honestly picks out 5
cards that i cannot see. he hands the five cards to me (joel can't see any of
this). i look at the cards and i pick 1 card out and give it back to babak. i
then arrange the other four cards in a special way, and give those 4 cards all
face down, and in a neat pile, to joel. joel looks at the 4 cards i gave him,
and says out loud which card babak is holding (suit and number).

i did not convey any information to joel other than the way i ordered the 4
cards, (all face down, aligned in line with one another) so how did i encode
babak's card using that method?

hint 1: card trick without the trick

hint 1: with five cards there will always be at least two cards of the same
suit. since i get to pick the card to give back to babak, i can use that to my
advantage to communicate the suit to joel.

hint 2: card trick without the trick

hint 2: if i communicate the suit to joel with one of the cards, then i'm only
left with the other 3 cards to show the number. since only the ordering of the
cards can be used, i am allowed up to six combinations. well, that won't do
to signify up to 13 cards, will it? certainly not, but if i pick the right card to
show the suit, i can use that card to reduce the range of cards down to 6.
how?
solution: card trick without a trick

this is a card trick without the trick. there is no sleight of hand, no tricks up
my sleeve, no magic whatsoever. it's all done with logic, yet it will amaze
most people.

joel and i are working together as a team to do the trick. babak will be the
culprit.

i ask babak to pick 5 cards out of a deck. he can pick any five cards, he can
shuffle the deck 7 times, it really doesn't matter. he honestly picks out 5
cards that i cannot see. he hands the five cards to me (joel can't see any of
this). i look at the cards and i pick 1 card out and give it back to babak. i
then arrange the other four cards in a special way, and give those 4 cards all
face down, and in a neat pile, to joel. joel looks at the 4 cards i gave him,
and says out loud which card babak is holding (suit and number).

i did not convey any information to joel other than the way i ordered the 4
cards, (all face down, aligned in line with one another) so how did i encode
babak's card using that method?

hint 1: card trick without the trick

hint 2: card trick without the trick

solution: hint 1 talks about how there are guaranteed to be at least 2 cards
with the same suit in a series of five cards. i use this to signal to joel the suit
of the card. i have to make a decision though between which of the two cards
to give back to babak. which one do i pick?

hint 2 talks about how if i use one of the cards to signal the suit, that only
leaves me with 3 cards. i can only arrange 3 cards in six permutations. how
do i signal a number between 2 and 13 using only six permutations?

simple really. consider that the series of cards is a cycle.

2 3 4 5 6 7 8 9 10 J Q K A 2 3 4 5 6 7 8 9 10 J Q K A 2 ... etc

i take the two cards of the same suit (if there are more than two of the same
suit, then i choose any two, it won't affect the solution) and i pass babak the
one that is further to the right in the cycle. if i do it correctly, the one further
to the right will be no more than 6 places away than the one to the left. (it
just depends on how you think about the cycle).

for example. say we had the 2 and the 9 of hearts. if you look at it as the 2
to the left and the 9 to the right, it is seven cards away. but if you look as
the 9 to the left and the 2 to the right in the cycle, the 2 is only 6 cards
away.

then i just come up with a system of the permutations that signals how many
places to the right to count to find the correct card

one such system could be:

sort the 3 non-same suit cards first by number then by suit (diamonds,
spades, hearts, clubs) to give a unique sort for every card.
label the cards as A, B, and C.
place the cards in the following order to signal how many places away
babak's card is:

A B C - 1

A C B - 2

B A C - 3

B C A - 4

C A B - 5

C B A - 6

then put the suit card on top and give the pile to joel.

example:
babak picks out the following five cards:
3C 7D 10S QH AD

the 7 of diamonds (7D) and the ace of diamonds (AD) are the two suit cards.
i recognize that AD is going to be the left card and 7D the right (because the
other way they are 7 spaces apart). i give babak back the 7D. i sort the other
cards, 3C 10S QH. i have to signal 6 places to the right (7D - AD = 6). so i
place them in C B A format: QH 10S 3C. i then place the AD on top and pass
to joel.

joel pulls off the first card and sees that it is the AD. (he knows babak's card
is a diamond). he looks at the next 3 cards and realizes their ordering is C B
A. signalling 6, he counts 6 spots from the ace. (2 3 4 5 6 7). then he says "7
of diamonds" and the crowd is amazed.

another sneakier card trick (with the trick) works as follows.

babak picks one card out of the deck and shows it to me. he doesn't tell joel
and i don't speak. after seeing the card, i sneakily use my one hand to signal
both the number and suit of the card. i use my four fingers (not the thumb)
as bits signalling a binary number. if they are extended it means 1, if i don't
extend them it means 0. the pinky is the least significant. so to signal 7 i
would extend all fingers except my index finger. the suit is signalled by four
positions of the thumb (hidden, adjacent to the hand, slightly ajar, fully
extended). i can use this method to tell joel the card without saying anything
(and making any noticeable motions). with some practice it can really amaze
people (at least it worked with my family).

bridge aha:!!

problem: this one is a classic that many of you have probably already heard,
but all the more reason why it should definitely be included here. four people
are on this side of the bridge. the bridge will be destroyed by a bomb in 17
minutes. everyone has to get across before that. problem is that it's dark and
so you can't cross the bridge without a flashlight, and they only have one
flashlight. plus the bridge is only big enough for two people to cross at once.
the four people walk at different speeds: one fella is so fast it only takes him
1 minute to cross the bridge, another 2 minutes, a third 5 minutes, the last it
takes 10 minutes to cross the bridge. when two people cross the bridge
together (sharing the flashlight), they both walk at the slower person's pace.
can they all get across before the bridge blows up?

person A: 1 minute
person B: 2 minutes
person C: 5 minutes
person D:10 minutes
solution: bridge

problem: four people are on this side of the bridge. the bridge will be
destroyed by a bomb in 17 minutes. everyone has to get across before
that. problem is that it's dark and so you can't cross the bridge without a
flashlight, and they only have one flashlight. plus the bridge is only big
enough for two people to cross at once. the four people walk at different
speeds: one fella is so fast it only takes him 1 minute to cross the bridge,
another 2 minutes, a third 5 minutes, the last it takes 10 minutes to cross
the bridge. when two people cross the bridge together (sharing the
flashlight), they both walk at the slower person's pace. can they all get
across before the bridge blows up?

person A: 1 minute

person B: 2 minutes

person C: 5 minutes

person D:10 minutes

solution: of course its possible, otherwise it wouldn't be a very interesting

question. the only trick is in realizing that you want to get the two slowest
people across together, because otherwise you are wasting too much time.
but then once you get them across, how do you not make one of them
walk back with the flashlight? well, you just have one of the fast people
already there waiting to sprint the flashlight back across.

1. A & B cross. total time: 2 minutes.

C |==========================| A

D | | B

|==========================| flashlight

2. B comes back. total time: 4 minutes.

C |==========================| A

D | |

B |==========================|

flashlight

3. C & D cross. total time: 14 minutes.

B |==========================| A

| | C

|==========================| D

flashlight

4. A comes back. total time: 15 minutes.

A |==========================| C

B | | D

|==========================|

flashlight

5. A & B cross. total time: 17 minutes.

 |========= =============| A

| KABOOM! | B

|========= =============| C D

flashlight

another solution which is valid is to have A bring the flashlight back in step
2. it only changes the solution slightly. this is supposed to be a "classic"
microsoft interview question but it seems absurdly easy to be a good
interview question (especially coupled with the fact that everyone has
probably heard it already).

jelly beans aha:!

you have three jars that are all mislabeled. one contains peanut butter jelly
beans, another grape jelly jelly beans, and the third has a mix of both (not
necessarily a 50/50 mix, could be a 1/99 mix or a 399/22 mix). how many
jelly beans would you have to pull out, and out of which jars, to find out how
to fix the labels on the jars?

| | | | | |
|jar 1| |jar 2| |jar 3|
| | | | | |
======= ======= =======
p.b. grape p.b./grape

solution: jellybeans

you have three jars that are all mislabeled. one contains peanut butter jelly
beans, another grape jelly jelly beans, and the third has a mix of both (not
necessarily a 50/50 mix, could be a 1/99 mix or a 399/22 mix). how many
jelly beans would you have to pull out, and out of which jars, to find out how
to fix the labels on the jars?

| | | | | |

|jar 1| |jar 2| |jar 3|

| | | | | |

======= ======= =======

p.b. grape p.b./grape

solution: 1 jelly bean from the p.b./grape jar will do the trick.

the trick here is to realize that every jar is mislabeled. therefore you know
that the peanut butter jelly bean jar is not the penut butter jelly bean jar,
and the same goes for the rest.

you also need to realize that it is the jar labeled p.b./grape, labelled as the
mix jar, that is your best hope. if you choose a jelly bean out of there, then
you will know whether that jar is peanut butter or grape jelly jelly beans. it
can't be the mix jar because i already said that every jar is mislabeled.

once you know that jar 3 is either peanut butter, or grape jelly, then you
know the other jars also. if it is peanut butter, then jar 2 must be mixed
because it can't be grape (as its labelled) and it can't be peanut butter (that's
jar 3). hence jar 1 is grape.

if jar 3 is grape, then you know jar 1 must be the mix because it can't be
p.b. (as its labelled) and it can't be grape (that's jar 3). hence jar 2 is peanut
butter.

if you pick jelly beans from jar 1 or jar 2, then you would have to pick out all
of the jelly beans before you knew what that jar was. this is because jar 1
and 2 could be the mix, so in order to disprove that they were the mix, you
would have to pull out every jelly bean just to make sure (since there could
just be one bean of the opposite flavor in there).
 bad king aha:!

a bad king has a cellar of 1000 bottles of delightful and very expensive
wine. a neighbouring queen plots to kill the bad king and sends a servant to
poison the wine. (un)fortunately the bad king's guards catch the servant
after he has only poisoned one bottle. alas, the guards don't know which
bottle but know that the poison is so strong that even if diluted 1,000,000
times it would still kill the king. furthermore, it takes one month to have an
effect. the bad king decides he will get some of the prisoners in his vast
dungeons to drink the wine. being a clever bad king he knows he needs to
murder no more than 10 prisoners - believing he can fob off such a low
death rate - and will still be able to drink the rest of the wine at his
anniversary party in 5 weeks time.

solution: bad king

problem: a bad king has a cellar of 1000 bottles of delightful and very
expensive wine. a neighbouring queen plots to kill the bad king and sends a
servant to poison the wine. (un)fortunately the bad king's guards catch the
servant after he has only poisoned one bottle. alas, the guards don't know
which bottle but know that the poison is so strong that even if diluted
1,000,000 times it would still kill the king. furthermore, it takes one month to
have an effect. the bad king decides he will get some of the prisoners in his
vast dungeons to drink the wine. being a clever bad king he knows he needs
to murder no more than 10 prisoners - believing he can fob off such a low
death rate - and will still be able to drink the rest of the wine at his
anniversary party in 5 weeks time.

explain how...

solution: i'll give you a hint. 1000 is less than 1024. if there were 1024 or
more bottles of wine it would take more than 10 prisoners.

number the bottles 1 to 1000, and write the number in binary format.
bottle 1 = 0000000001

bottle 250 = 0011111010

bottle 1000 = 1111101000

now take your prisoner's 1 through 10 and let prisoner 1 take a sip from
every bottle that has a 1 in its least significant bit. let prisoner 10 take a sip
from every bottle with a 1 in its most significant bit. etc.

prisoner 10 9 8 7 6 5 4 3 2 1

bottle 924 1 1 1 0 0 1 1 1 0 0

for instance, bottle #924 would be sipped by 10,9,8,5,4 and 3. that way if
bottle #924 was the poisoned one, only those prisoners would die.

after four weeks, line the prisoners up in their bit order and read each living
prisoner as a 0 bit and each dead prisoner as a 1 bit. the number that you
get is the bottle of wine that was poisoned.

additional question: to increase your chance of living, which prisoner would

you want to be?

if there were 1023 bottles, it wouldn't matter since everyone would have to
take 512 sips. but there are 23 bottles less, so the people whose bits would
have been on from 1001 to 1023 won't have to take a sip. 1001 is [11111
01001] in binary and 1023 is [11111 11111]. the most five significant bits
are the most interesting because they would always be on from 1001 to
1023, so all those people are missing out on 23 bottles of wine that they
otherwise would have had to drink. so in order to increase your chance of
living, you'd probably want to be prisoner 6 to 10. (but depending on how
the king determines who is least significant and who is most significant you
could get shafted.)

note that if the king was really trying to kill the least number of prisoners, he
should have let 999 prisoners each take a sip from their respective bottle
numerically (if he had that many prisoners at his disposal). that way only one
prisoner would die, and there's a chance of 1/1000 that no one would die,
but then the puzzle isn't very fun.

fogcreek programmers aha:!

100 fogcreek programmers are lined up in a row by an assassin. the assassin

puts red and blue hats on them. they can't see their own hats, but they can
see the hats of the people in front of them. the assassin starts in the back
and says "what color is your hat?" the fogcreek programmer can only answer
"red" or "blue." the programmer is killed if he gives the wrong answer; then
the assassin moves on to the next programmer. the programmers in front
get to hear the answers of the programmers behind them, but not whether
they live or die. they can consult and agree on a strategy before being lined
up, but after being lined up and having the hats put on, they can't
communicate in any way other than those already specified. what strategy
should they choose to maximize the number of programmers who are
guaranteed to be saved?

solution: fog creek programmers

problem: 100 fogcreek programmers are lined up in a row by an assassin.

the assassin puts red and blue hats on them. they can't see their own hats,
but they can see the hats of the people in front of them. the assassin starts
in the back and says "what color is your hat?" the fogcreek programmer can
only answer "red" or "blue." the programmer is killed if he gives the wrong
answer; then the assassin moves on to the next programmer. the
programmers in front get to hear the answers of the programmers behind
them, but not whether they live or die. they can consult and agree on a
strategy before being lined up, but after being lined up and having the hats
put on, they can't communicate in any way other than those already
specified. what strategy should they choose to maximize the number of
programmers who are guaranteed to be saved?
this is a very difficult problem to solve during an interview (especially if
you've already taxed the candidate's brain). look for obvious solutions first,
and the reasoning behind them and then try to lead them to the ultimate
solution.

a logical answer could be all the programmers would just say "red" and that
way about half of them would survive on average, assuming the hats were
distributed randomly.

this is a good start and should naturally lead to having every other
programmer say the color of the hat in front of them. the first programmer
would say the color of the hat in front of him, then the next programmer
would just say that color that was just said. so we can guarantee that half
survive - the even numbered programmers (since the person behind them
told them the answer). and potentially if the hats were distributed randomly
some of the programmers would get lucky and the hat in front of them would
be the same color as their own. so this strategy should save more than half,
and on average 75% of them would live.

at this point, if the solution is not clear, the candidate may give answers like,
"they could agree that if they said their hat color in a soft voice, it means the
hat in front of them is the same color, and if they say it in a loud voice, it
means the hat in front is a different color". this is definitely good and on the
correct track. another option is they could say "reeeeeeeeeeed" for x number
of seconds, where x represented the distribution of hats where a hat was a
bit in a binary number, (red = 1, blue = 0). another interesting answer.
there are many others like these that "bend" the rules and come to a
solution.

but the real solution acknowledges that the programmers can only say "red"
or "blue" and cannot alter their voice in such a convincing way as to signal
any information other than the word they said. a good way to get this point
across, is simply to change the problem slightly by saying "the assassin gets
to hear their plan before she puts the hats on, and so will try to thwart the
plan however she can."

so if they decide to all say "red", she'll put blue hats on all of them. if they
decide to all say the color of the hat in front of them, she'll alternate the hats
on every head, guaranteeing half will die. even with the assassin hearing
their plan, there is still a way to save almost everyone.

we know that the first person is never going to have any information about
the color of their hat, so they cannot be guaranteed to survive. but, i'll give
you a hint to the solution: i can save every other person for sure.

solution: they agree that if the number of red hats that the back person can
see is even, that programmer will say "red". if they add up to an odd
number, they will say "blue". this way number 99 can look ahead and count
the red hats. if they add up to an even number and number 100 said "red",
then 99 must be wearing a blue hat. if they add up to an even number and
number 100 said "blue", signalling an odd number of red hats, number 99
must also be wearing a red hat. number 98 knows that 99 said the correct
hat, and so uses that information along with the 97 hats in front to figure out
what color hat is on 98's head.

sample:

100 99 98 97 96 95 94 ... facing ->

R B B R B R B ... -> 45 R and 48 B

this shows #100 wearing a red hat, 99 a blue, 98 a blue, 97 a red, 96 a blue,
95 a red, 94 a blue and 45 red hats - 48 blue hats on the people in front of
them.

100 counts up the red hats: 47 total. so 100 says "blue". the assassin kills
100. 99 counts up the red hats in front: 47. 100 said blue, so 100 saw an
odd number. 99 sees an odd number, so 99 says "blue" and lives. 98 had
counted 47 red hats, and 99 didn't say "red" so thats still the total. 98 says
"blue". 97 counts up and finds 46 red hats. 99 and 98 didn't say "red", so his
count is missing a red hat (its on his head, he realizes). he says "red". 96
heard the "red" and now knows that there are an even number of "red" hats
in front of 95. 96 sees 46, so he knows he has a "blue" hat. etc...

even if the assassin knows the plan, she can't thwart it. she hears the plan,
but she still has to put the hats on their heads. the plan doesn't rely on any
ordering of the hats, so the worst the assassin can do is to make sure #100
gets killed and thats the worst damage she can do.

pirates aha:!

five pirates have 100 gold coins. they have to divide up the loot. in order of
seniority (suppose pirate 5 is most senior, pirate 1 is least senior), the most
senior pirate proposes a distribution of the loot. they vote and if at least 50%
accept the proposal, the loot is divided as proposed. otherwise the most
senior pirate is executed, and they start over again with the next senior
pirate. what solution does the most senior pirate propose? assume they are
very intelligent and extremely greedy (and that they would prefer not to die).

solution: pirates

five pirates have 100 gold coins. they have to divide up the loot. in order of
seniority (suppose pirate 5 is most senior, pirate 1 is least senior), the most
senior pirate proposes a distribution of the loot. they vote and if at least 50%
accept the proposal, the loot is divided as proposed. otherwise the most
senior pirate is executed, and they start over again with the next senior
pirate. what solution does the most senior pirate propose? assume they are
very intelligent and extremely greedy (and that they would prefer not to die).

(to be clear on what 50% means, 3 pirates must vote for the proposal when
there are 5 for it to pass. 2 if there are 4. 2 if there are 3. etc... )
solution: most of the time i get people who give answers like "the most
senior pirate takes half and divides the rest up among the least senior
pirates." um, you missed the whole point to begin with. sorry.

any answer without a specific logic behind it is invalid. if i ask you why pirate
5 gave x coins to pirate 1, please don't say "because he's nice".

now for the real solution. pirate 5 being the most senior knows that he needs
to get 2 other people to vote for his solution in order for him not to be
executed. so who can he get to vote for him, and why would they choose to
vote for him? if you start thinking that pirate 4 will never vote for him,
because he would rather have 5 die and then be in charge and take it all for
himself, you are on the right track. but it gets more complicated.

lets consider if there were only 1 pirate. obviously he would take it all for
himself and no one would complain.

if there were 2 pirates, pirate 2 being the most senior, he would just vote for
himself and that would be 50% of the vote, so he's obviously going to keep
all the money for himself.

if there were 3 pirates, pirate 3 has to convince at least one other person to
join in his plan. so who can he convince and how? here is the leap that needs
to be made to solve this problem. pirate 3 realizes that if his plan is not
adopted he will be executed and they will be left with 2 pirates. he
already knows what happens when there are 2 pirates as we just figured out.
pirate 2 takes all the money himself and gives nothing to pirate 1. so pirate 3
proposes that he will take 99 gold coins and give 1 coin to pirate 1. pirate 1
says, well, 1 is better than none, and since i know if i don't vote for pirate 3,
i get nothing, i should vote for this plan.

now we know what happens when there are 3 pirates. so what happens with
4? well pirate 4 has to convince 1 other person to join in his plan. he knows if
he walks the plank then pirate 3 will get 99 coins and pirate 1 will get 1 coin.
pirate 4 could propose giving pirate 1 two coins, and surely pirate 1 would
vote for him, since 2 is better than 1. but as greedy as he is, pirate 4 would
rather not part with 2 whole coins. he realizes that if he gets executed, then
pirate 3's scenario happens and pirate 2 gets the shaft in that scenario (he
gets zero coins). so pirate 4 proposes that he will give 1 coin to pirate 2, and
pirate 2 seeing that 1 is better than 0 will obviously vote for this plan.

a common objection is that pirate 2 is not guaranteed to vote for this plan
since he might hope for the case when there are only 2 pirates and then he
gets all the booty. but that is why i said that the pirates are extremely
intelligent. pirate 2 realizes that pirate 3 is smart enough to make the
optimal proposal, so he realizes that there will never be 2 pirates left,
because 3 doesn't want to die and we just showed that 3 has a winning
proposal.

so lets sum up at this point

Pirate 1 2 3 4 5

5. ? ? ? ? ?

4. 0 1 0 99 -

3. 1 0 99 - -

2. 0 100 - - -

1.100

once you see the pattern it becomes very clear. you have to realize that
when a pirate's plan does not succeed then that means you are in the same
situation with one less pirate.
1. pirate 1 needs 0 other people to vote for him. so he votes for himself and
takes all the money. 2. pirate 2 needs 0 other people to vote for him. so he
votes for himself and takes all the money. pirate 1 gets 0. 3. pirate 3 needs 1
other person to vote for him. he gives 1 coin to pirate 1 for his vote - if we
are reduced to 2 pirates, pirate 1 gets 0 so pirate 1 knows 1 is better than
none. pirate 3 takes 99. pirate 2 gets 0. 4. pirate 4 needs 1 other person to
vote for him. he gives 1 coin to pirate 2 - if we reduce to 3 pirates, pirate 2
gets 0 so pirate 2 knows 1 is better than none. pirate 4 takes 99. pirate 3
gets 0. pirate 1 gets 0. 5. pirate 5 needs 2 other people to vote for him. its
clear now that the 2 people he needs to convince are the 2 who get shafted
in the 4 pirate scenario - pirate 3 and pirate 1. so he can give them each 1
coin (which is better than 0 - what they would get otherwise) and keep 98
for himself.

Pirate 1 2 3 4 5

5. 1 0 1 0 98

what happens if there are 15 pirates? pirate 15 needs 7 other people to vote
for him, so he recruits pirates 13,11,9,7,5,3, and 1 with 1 coin each and
keeps 93 coins himself. those pirates will all vote for him because they know
that they get 0 coins if he dies and pirate 14 is in charge.

hope you enjoyed this one. its my favorite interview question of all. it really
allows the candidate to ask a lot of interesting questions and its really
amazing when they reach the solution all by themselves (as all fogcreek
employees have done so far).

sum it up aha:!

problem: you are given a sequence of numbers from 1 to n-1 with one of the
numbers repeating only once. (example: 1 2 3 3 4 5). how can you find the
repeating number? what if i give you the constraint that you can't use a
dynamic amount of memory (i.e. the amount of memory you use can't be
related to n)?
what if there are two repeating numbers (and the same memory constraint?)

solution: sum it up'

problem: you are given a sequence of numbers from 1 to n-1 with one of the
numbers repeating once. (example: 1 2 3 3 4 5) how can you find the
repeating number?

solution:

as a programmer, my first answer to this problem would be make a bit

vector of size n, and every time you see the number, set its correspond index
bit to 1. if the bit is already set, then that's the repeater. since there were no
constraints in the question, this is an ok answer. its good because it makes
sense if you draw it for someone, whether they are a programmer,
mathemetician, or just your grandpa. its not the most efficient answer
though.

now, if i add the constraint that you can only use a fixed amount of memory
(i.e. not determined by n) and it must run in O(n) time... how do we solve it.
adding all the numbers up from 1 to n-1 would give us a distinct sum.
subtracting the total sum of all the numbers from the sum of n to n-1 ( which
is (n)(n-1)/2 ) would give us the secret extra number.

what if you can only use a fixed amount of memory, and two of the numbers
are repeated? we know that the numbers have a distinct sum, and the
difference would be equal to the sum of our unknowns
c = a + b

where c is the sum and a and b are the unknowns - c is a constant

if we had another similar formula we could solve the two unknown equations.
my first thought was that the numbers would have a distinct product - (n-1)!
if we divide the total product by the (n-1)! product, we would get another
equation
c2 = ab

we could then solve the two equations to get them into quadratic formula
notation
0 = ax^2 + bx + c
and solve for the two values of x. this answer is correct but factorial grows
really fast.

some sort of sum would be better. the sum of the squares from n-1 to 1
would work. that would yield a function of the form
c2 = a^2 + b^2

which could also be solved by using the quadratic equation.

i think its fine to remind someone of the quadratic equation... (maybe only
because i myself had to look it up to solve the problem) i mean really
though, the last time i used it was probably in 10th grade. as long as they
get the idea that given two unknowns and two equations you can solve for
the unknowns - thats the point.

palindromes aha:!!

problem: this year on October 2, 2001, the date in MMDDYYYY format will be
a palindrome (same forwards as backwards).
10/02/2001
when was the last date that this occurred on? (see if you can do it in your
head!)

solution: palindrome

problem: this year on October 2, 2001, the date in MMDDYYYY format will be
a palindrome (same forwards as backwards).
10/02/2001
when was the last date that this occurred on? (see if you can do it in your
head!)
solution: we know the year has to be less than 2001 since we already have
the palindrome for 10/02. it can't be any year in 1900 because that would
result in a day of 91. same for 1800 down to 1400. it could be a year in 1300
because that would be the 31st day. so whats the latest year in 1300 that
would make a month? at first i thought it would be 1321, since that would
give us the 12th month, but we have to remember that we want the
maximum year in the 1300 century with a valid month, which would actually
be 1390, since 09/31 is a valid date.

but of course, a question like this wouldn't be complete without an aha

factor. and of course, there are not 31 days in september, only 30. so we
have to go back to august 08 which means the correct date would be
08/31/1380.

palindromes also offer another great string question.

write a function that tests for palindromes
bool isPalindrome( char* pStr )

if you start a pointer at the beginning and the end of the string and keep
comparing characters while moving the pointers closer together, you can test
if the string is the same forwards and backwards. notice that the pointers
only have to travel to the middle, not all the way to the other end (to reduce
redundancy).

bool isPalindrome( char* pStr )

{
if ( pStr == NULL )
return false;

char* pEnd = pStr;

while ( *pEnd != '\0' )
pEnd++;
 pEnd--;

while(pEnd > pStr)

{
if ( *pEnd != *pStr )
return false;

pEnd--;
pStr++;
}

return true;
}

daughter's ages aha:!!

two MIT math grads bump into each other at Fairway on the upper west side.
they haven't seen each other in over 20 years.

the first grad says to the second: "how have you been?"
second: "great! i got married and i have three daughters now"
first: "really? how old are they?"
second: "well, the product of their ages is 72, and the sum of their ages is
the same as the number on that building over there.."
first: "right, ok.. oh wait.. hmm, i still don't know"
second: "oh sorry, the oldest one just started to play the piano"
first: "wonderful! my oldest is the same age!"

problem: how old are the daughters?

solution: daughter's ages

two MIT math grads bump into each other at Fairway on the upper west side.
they haven't seen each other in over 20 years.
the first grad says to the second: "how have you been?"
second: "great! i got married and i have three daughters now"
first: "really? how old are they?"
second: "well, the product of their ages is 72, and the sum of their ages is
the same as the number on that building over there.."
first: "right, ok.. oh wait.. hmm, i still don't know"
second: "oh sorry, the oldest one just started to play the piano"
first: "wonderful! my oldest is the same age!"

problem: how old are the daughters?

solution: start with what you know. you know there are 3 daughters whose
ages multiply to 72. let's look at the possibilities...

Ages: Sum of ages:

1 1 72 74

1 2 36 39

1 3 24 28

1 4 18 23

1 6 12 19

1 8 9 18

2 2 18 22

2 3 12 17

2 4 9 15

2 6 6 14

3 3 8 14

3 4 6 13

after looking at the building number the man still can't figure out what their
ages are (we're assuming since he's an MIT math grad, he can factor 72 and
add up the sums), so the building number must be 14, since that is the only
sum that has more than one possibility.
finally the man discovers that there is an oldest daughter. that rules out the
"2 6 6" possibility since the two oldest would be twins. therefore, the
daughters ages must be "3 3 8".
(caveat: an astute reader pointed out that it is possible for two
siblings to have the same age but not be twins, for instance one is
born in january, and the next is conceived right away and delivered in
october. next october both siblings will be one year old. if a candidate
points this out, extra credit points to him/her.)

this question is pretty neat, although there is certainly a bit of an aha factor
to it. the clues are given in such a way that you think you are missing
information (the building number), but whats important isn't the building
number, but the fact that the first man thought that it was enough
information, but actually wasn't.

even if the candidate doesn't know the solution, they could come up with
some interesting thoughts. if they just stare at you and shrug "i dunno" then
thank them for their time and don't give them a fogcreek pen.

int atoi( char* pStr ) aha:!

problem: write the definition for this function without using any built-in
functions. if pStr is null, return 0. if pStr contains non-numeric characters,
either return 0 (ok) or return the number derived so far (better) (e.g. if its
"123A", then return 123). assume all numbers are positive. plus or minus
signs can be considered non-numeric characters. in order to solve this
program, the programmer must understand the difference between the
integer 0 and the character '0', and how converting '0' to an int, will not
result in 0. in other words, they have to understand what ascii is all about.

solution: atoi

string manipulation functions are great programming questions. they test

whether the user can understand and translate into code simple algorithms.
string functions test pointer arithmetic which usually shows a knowledgeable
programmer. also there are usually multiple solutions, some more efficient
than others. plus people use them all the time so they should understand
how they work. my favorite is atoi and i start the problem like this:

int atoi( char* pStr )

write the definition for this function without using any built-in functions. if
pStr is null, return 0. if pStr contains non-numeric characters, either return 0
(ok) or return the number derived so far (better) (e.g. if its "123A", then
return 123). assume all numbers are positive. plus or minus signs can be
considered non-numeric characters. in order to solve this program, the
programmer must understand the difference between the integer 0 and the
character '0', and how converting '0' to an int, will not result in 0. in other
words, they have to understand what ascii is all about. if they are stuck
solving this problem, just ask them first to write:

charToInt(char c)

if they can't do that then they basically missed half the problem. any
moderately talented programmer who has a CS degree knows how to convert
a char to an int. (note i said convert, not cast. charToInt('9') should return
9.)

when they start to solve the problem you will notice that they must make a
choice in how they will process the string - from left to right or right to left. i
will discuss both methods and the difficulties encountered in each.

"right to left" - this method starts at the right hand letter of the string and
converts that character to an int. it then stores this value after promoting it
to its correct "tens" place.

int atoi( char* pStr )

{
 int iRetVal = 0;
int iTens = 1;

if ( pStr )
{
char* pCur = pStr;
while (*pCur)
pCur++;

pCur--;

while ( pCur >= pStr && *pCur <= '9' && *pCur >= '0' )
{
iRetVal += ((*pCur - '0') * iTens);
pCur--;
iTens *= 10;
}
}
return iRetVal;
}

"left to right" - this method keeps adding the number and multiplying the
result by ten before continuing to the next number. e.g. if you had "6234"
and you processed from left to right you'd have 6, then if you kept reading
you'd multiply your result by 10 (6*10) to add a zero for where the next
number would go. 60, and then you'd slide the 2 into the zero place you just
made. 62. do it again, 620, slide the next number in, 623.

int atoi( char* pStr )

{
int iRetVal = 0;

if ( pStr )
 {
while ( *pStr && *pStr <= '9' && *pStr >= '0' )
{
iRetVal = (iRetVal * 10) + (*pStr - '0');
pStr++;
}
}
return iRetVal;
}

i think the "left to right" method is a little bit cleaner, or maybe its just
cooler. but both are "correct".

remember that debugging code on paper is somewhat hard. most

programmers aren't used to studying code that much when you can just hit
F-7, compile and see if the compiler barfs or not. if you notice an error, just
ask them to step through a sample string drawing out what is happening with
all the variables and the pointers in every step. they should find their mistake
then and fix it (no points deducted).

bumblebee aha:!!!!

problem: two trains enter a tunnel 200 miles long (yeah, its a big tunnel)
travelling at 100 mph at the same time from opposite directions. as soon as
they enter the tunnel a supersonic bee flying at 1000 mph starts from one
train and heads toward the other one. as soon as it reaches the other one it
turns around and heads back toward the first, going back and forth between
the trains until the trains collide in a fiery explosion in the middle of the
tunnel (the bee survives). how far did the bee travel?

solution: bumblebee

problem: two trains enter a tunnel 200 miles long (yeah, its a big tunnel)
travelling at 100 mph at the same time from opposite directions. as soon as
they enter the tunnel a supersonic bee flying at 1000 mph starts from one
train and heads toward the other one. as soon as it reaches the other one it
turns around and heads back toward the first, going back and forth between
the trains until the trains collide in a fiery explosion in the middle of the
tunnel (the bee survives). how far did the bee travel?

solution: this puzzle falls pretty high on my aha scale. my first inclination
when i heard it was to think "ok, so i just need to sum up the distances that
the bee travels..." but then you quickly realize that its a difficult (not
impossible) summation which the interviewer could hardly expect you to
answer (unless i guess if you are looking for a job as a quant). "there must
be a trick" you say. eh, sort of i guess, enough to say that this question is a
stupid interview question.

the tunnel is 200 miles long. the trains meet in the middle travelling at 100
mph, so it takes them an hour to reach the middle. the bee is travelling 1000
mph for an hour (since its flying the whole time the trains are racing toward
one another) - so basically the bee goes 1000 miles.

there is no process to explain, so this question can't possibly teach you

anything about the person. they either know it or they don't and if they
already knew it before you asked, you're not going to be able to tell when
they give you the answer. so don't ask this question. and if someone asks
you this question, just tell them you've already heard it before.