puzzles & curiosities

dave bogle


SITE SEARCH
HOME
ARCHIVES

PERSONAL
MY CV
SCHOOL PHOTOS

ARTS
HOME PAGE
COPYRIGHT LAW
60'S CHARTS
70'S CHARTS
80'S CHARTS
ARTIST OF MONTH
ART ON THE WEB

CHESS
ENDGAME PROGRAM
CHESS ON THE WEB

SCRIPTS
DOMINOES
YANKEES ETC

MATHS & SCIENCE
HOME PAGE
MEDICAL
PUZZLES
MYTHS
HANDS OF BRAG
HANDS OF POKER





© dw bogle 2001

 
maths & science odds of poker odds of brag medical myths
 
 22 Apr 2002
 

What is the most common digit?
Daft as it may sound, there is indeed such a thing, and the digit is 1.  It doesn't matter whether you're talking about the lengths of rivers (in miles, kilometres, cubits or any other measurement), the populations of cities, the item prices on your shopping receipts or whatever: in each random set of numbers 1 is the most common of all digits.  This is Benford's Law, which is now understood to a sophisticated extent, and is used by the United States Inland Revenue Service and others to check for non-standard (i.e. doctored) distributions on corporate returns etc.

How many ways can you arrange things?
For 3 things, you can put them in 1 × 2 × 3 different orders, like this:

  • 1-2-3
  • 1-3-2
  • 2-1-3
  • 2-3-1
  • 3-1-2
  • 3-2-1

The formula is 1 × 2 × 3 × 4 for four objects, and so on.  These are factorials, denoted by an exclamation mark e.g. factorial 4 = 4! 

If you're putting individual letters into addressed envelopes, and you wickedly want to know how many different ways you can put all the letters into the wrong envelopes, this is far more involved.   Just one example: with 5 letters, there are 44 ways of getting every single one wrong, leaving 75 ways of getting some of them wrong, and only one way of getting the job right. 

How do you calculate Pi?
Here is a routine in pseudo-code:

Initial Values of Variables
a = 1; d = 1; y = 1; z = 1;
b = ( 1 / sqrt(2) )
c = 0.25
Loop indefinitely
y = a
a = ( a + b ) / 2
b =  sqrt(b × y)
c = c - (z × (a - y)2 )
z = z × 2
display (a + b)2 / (c × 4)

How many colours are needed for a map?
Four.  This has been known since 1878, but was not proved until 1976.  The final proof was hundreds of pages long, took 1200 computer-hours of calculations, and is beyond the understanding of all but a handful of specialist mathematicians.

How do you construct a magic square?

12 13 1 8
6 3 15 10
7 2 14 11
9 16 4 5
The rows, columns and diagonals all add up to 34.

What is the largest number?
A googol is 10100 which is pretty big,  and a googolplex is 10googol, which is even larger.  But the winner by billions of universes is Graham's Number, which dwarfs both of these, and for that matter any other quantity ever conceived.  Graham's number is the answer to a problem in Combinatorics (an abstruse branch of Maths dealing with sets) and although the algorithm for calculating it can be described, the number itself is so large that it cannot be written down in any existing notation at all - even using towers of exponents.  Why not?  Because there would not be enough material in the universe to do so.  This is a monster of a number. 

 
PREVIOUS PUZZLES
The Nov-Dec puzzle was: 
Find the two numbers, the sum of which is 10, and the product of which is 40.  Or put another way, what are the values of x and y when (x+y)=10 and xy=40?  There is no trickery with bases: 10 = Ten and 40 = Forty.
The solution is that one number =
5 + sqrt(-15) and the other number =
5 - sqrt(-15).   This should be enough for you, provided you remember that:
(a + b) * ( a - b ) = (a * a) - (b * b)

If anyone's still lost, then please email me, and I'll take you through it.

The Jan-Feb puzzle asked you to give the next item in this series:

1, 4, 7, 11, 15, 18, 21, 24, 27...

A bit of lateral thinking required here, as it was not really a Maths puzzle at all.  The solution lies in the number of letters in the English names for these numbers.  73 (SEVENTYTHREE) is the answer 

BOTTLE IMP PARADOX

I had quite a few emails from people who did not understand this problem.  Here it is.

There is a short story by Robert Louis Stevenson called The Bottle Imp, which centres on an imp who will grant its owner their every wish - no strings attached.   Ask for £1bn and a lovestruck Meg Ryan, and it's all yours for keeps.   But if you're still in possession of the imp when you die, you'll spend an eternity in Hell.  Other thing is that the only way to renounce ownership of the imp is to sell it on to someone for less than you paid for it, so the question is: 

What is the minimum price that it is sensible to pay for the imp? 

Right, so you're not going to buy it for 1p, since you can never get rid of it.  Nor will you buy it for 2p, since no-one is going to buy it off you for 1p.   But if it's illogcal to buy it for 2p, then it must be illogical to buy it for 3p, since no-one will buy it for 2p.

See where we're going?   In a world of logicians there would simply be no price at all that it would be sensible to pay for the imp.   Now in practice there will indeed be a price at which you might reckon people will not follow their logic through to the end.  But do it at your own risk!!


YOU MISSED YOUR CHANCE!
Until 15 Mar 2002 Faber & Faber were offering a nice little earner ($1million) to anyone who could prove Goldbach's Conjecture:

In 1742, Christian Goldbach speculated that every even number greater than 2 could be expressed as the sum of two primes.  This has been verified for all numbers up to about 400 trillion, but it has never actually been proved.  All you needed to do is send Faber & Faber the proof before 15 Mar 2002, and you'd have been home and dry.   Hard luck.

PI IN THE SKY
The history of Pi is littered with delusions and false starts.  In 1853 William Shanks published his calculations of Pi to 707 decimal places.  How long that took him by hand you wouldn't want to think about, but they had determination in those days.  In 1945, it was discovered that Shanks had made an error in his calculation of the 528th place, thus making his last 180 digits wrong.  Mercifully, he'd been dead for 63 years.
More fun and games in 1857, when Edwin J Goodwin inspired the introduction of a bill in the House of Representatives, to enshrine in American law a supposedly exact value of Pi.  This optimistic bill got through its first reading, but was then held up on the intervention of a mathematician who by chance was visiting the House.  Its second reading is still to be held.

PRIMES AGAIN
Not all conjectures have lasted as long as Goldbach's.  Pierre de Fermat (1601-1665) thought he had discovered a formula for generating primes with:

 [2^(2^n)] +1

where ^ indicates exponentiation.   In 1742, however, Euler discovered this to be false where n=5.  To be fair to Fermat (one of the greatest mathematicians of all time) this is the only one of his many conjectures ever to have been proved wrong.



A BAD CONJECTURE
No such plaudits for de Polignac, who conjectured that every odd number was the sum of a power of 2 and a prime.   De Polignac claimed to have verified this up to 3 million, and if he did indeed go to these lengths it was a monumental waste of effort;  for if he had only been more careful when examining the number 127, for example...