home | index | units | counting | geometry | algebra | trigonometry & functions | calculus
analysis | sets & logic | number theory | recreational | misc | nomenclature & history | physics
 Blaise Pascal 
 1623-1662 Christiaan Huygens 
 1629-1695 Jacques Bernoulli 
 1655-1705 Leibniz

Final Answers
© 2000-2019   Gérard P. Michon, Ph.D.

Combinatorics and Probabilities


Related articles on this site:

Articles formerly on this page:

Related Links / Further Reading (Outside this Site)

The Electronic Journal of Combinatorics.
AOL: Combinatorics.
Combinatorics and more...  Blog of  Gil Kalai  (b. 1955).
Probability, symmetry, linearity  [1, 2, 3, 4, 5, 6Mikhail Gromov  (IHES, 2014).

Combinatorics & Probability     [Utility]

(2002-02-05) The Monty Hall Paradox
In a game-show, the contestant wins if he guesses correctly which one of three doors hides the (only) prize.  The rules of the game state that the contestant makes a tentative guess and is then shown one wrong choice among the two doors which he did not pick.  He is then given the opportunity to change his guess.  Should he do so?

Yes, always!  The chances of winning are doubled by switching this way.

The reason is that the first guess is right with probability 1/3, whereas one of the other two choices is correct with probability 2/3. The rules are actually a disguise for a simple choice between two alternatives: Would you rather win if the prize is behind the door you first picked (1/3) or win if it's not (2/3)?

Monty Hall, 1975 Before making national news in 1991, as chronicled below, this puzzle appeared under several other guises,  including  Bertrand's Box Paradox  introduced by Joseph Bertrand in  Calcul des probabilités (1889)  and The Three Prisoner Problem, which Martin Gardner presented in 1959, in his Scientific American column (and again in his 1961 book More Mathematical Puzzles).  The popularity of Monty Hall's TV gameshow Let's Make a Deal revived interest in the problem, about which Steve Selvin wrote a letter to the editor entitled "A Problem in Probability", published in the February 1975 issue of The American Statistician (with a follow-up in August 1975). Marilyn vos Savant

Most notably, the problem was posed by a reader of the famous puzzle columnist Marilyn vos Savant (Parade Magazine, September 9, 1990), whose correct answer started a flood of objections.  The thing became known as the Monty Hall Problem, in honor of the popular host of the TV show Let's Make a Deal (which ran 4500 times from 1963 to 1990).  The craze culminated in a front-page article of the  Sunday New York Times,  on July 21, 1991.  A number of people just could not believe the above answer was correct (they were wrong).  Fewer people remarked that the above rules were not directly applicable to the situation in Let's Make a Deal (they are right).  Actually, it's quite interesting to investigate how different "rules" affect the conclusion:

What if the show's host is not obligated to reveal a wrong choice?

We can't begin to address the issue if we don't have a clue about the host's obligations. For example, the producers may ask him to flip a fair coin in order to decide whether or not to proceed as above (in such a case the contestant should still take the opportunity to switch whenever available, but his overall probability of winning is reduced from 2/3 to 1/2).  It may also be known to the contestant that the "hostile" host only gives the opportunity to switch when the initial guess was right (in which case the contestant should clearly always decline any such "opportunity" and will win with probability 1/3).  A generalized analysis may be carried out along the following lines:

From watching many previous shows, the contestant knows that he will be given an opportunity to switch with probability  x  if his original guess was right and with probability  y  otherwise. 
Should he switch when given the opportunity to do so?

A contestant who  always  switches will win with probability (1-x)/3 + 2y/3 .  The first term is the probability the contestant was right and can't switch, whereas the second term is the probability he was wrong and could switch. It's better to switch whenever that probability is larger than 1/3  (which is the probability of winning when one never switches).  In other words, a contestant  should  switch whenever 2y is at least equal to x.

A host may want to keep his show interesting by having (clever) contestants react differently (or indifferently) to the switching opportunity.  The above demonstrates that such a host should allow good guesses to be changed twice as often as he does for bad guesses (so that x = 2y).  This makes the contestant win with probability 1/3, whether he decides to switch or not.

(2002-04-01)   The Three Prisoners Paradox  [abridged]
Of three prisoners (A, B and C), only the one secretly pardoned by the governor will escape execution. The governor picked at random and told the warden. The warden refuses to tell the fate of A, but agrees to name another prisoner who's doomed; he reveals that B was not pardoned. The survivor will thus be A or C. What's the probability that it will be A?

Before the warden's revelation, the probability that A would survive was clearly 1/3, as the governor's "random" choice was presumably not biased in favor of anyone. The confirmation of the execution of either one of the other two prisoners does not changes that. It's irrelevant to the fate of A, whose probability of survival remains 1/3.

However, things are looking brighter for C who is now known to survive with probability 2/3, since he escapes execution whenever A does not.

This is another version of the infamous "paradox" discussed in the previous article. However, this presentation of the so-called Three Prisoners Paradox predates the equivalent Monty Hall Paradox by several decades, as it was introduced by Martin Gardner in the October 1959 issue of Scientific American.  A recent account may be found in Gardner's own 2001 book: The Colossal Book of Mathematics, W.W. Norton & Company, New York, ISBN 0-393-02023-1.

( J. S. of Parkville, MD. 2000-05-03)
In how many ways can 6 children be seated at a circular table?

Answer :   120 ways.

This is the factorial of 5  (namely, 5! = 5´4´3´2´1); the number of ways the 5 other children may be placed to the left of some arbitrarily chosen kid.

5 ways to choose the kid placed first to the left, 4 ways to chose the second, 3 ways for the third, 2 ways for the fourth, 1 way for the last...

Bachet Square (2002-02-15)   1152
Bachet square  is a 4 by 4 layout of the 16 court cards (aces and faces) where every suit and every value appears only once in every row or column and in either diagonal.
How many different  Bachet squares  are there ?

Given a first row of 4 distinct letters, only two squares exist where each letter appears once in each row, in each column, and in either diagonal:

D   D

There are  576 = (4!)2  choices for the first row in a Bachet square  (4! = 24  ways to choose the order of the suits and 4! = 24 choices for the values).  Given the first row, we obtain two Bachet squares, (using either the first pattern for suits and the second one for values, or vice versa).  All told:  1152  Bachet squares.

 Bachet de Meziriac 
   Claude-Gaspard Bachet 
 1581-1638 Thus, among the  16! = 20922789888000  possible squares of the court cards, only one in  18162144000  is a  Bachet squareClaude-Gaspard Bachet de Méziriac (1581-1638) described those rare arrangements in 1624.  They resurfaced in Récréations mathématiques et physiques  by Jacques Ozanam (1694).
Claude-Gaspard Bachet was the first occupant (1634) of the 13th seat at the  Académie Française,  although he did not attend the founding ceremonies, on account of ill health.  The modern records of the Academy alphabetize him as  "Méziriac (Bachet de)".  Here are a few notorious occupants of his seat:  Jean Racine (1672), Octave Feuillet (1862), Pierre Loti (1891), Paul Claudel (1946), Maurice Schumann (1974), Pierre Messmer (1999) and Simone Veil (2008).

(2000-10-08, by popular request)   An Introduction to  Choice Numbers
What does C(n,p) mean?   [ We don't recommend the notation  nCp. ]

C(n,p) is usually pronounced "n choose p".  It is simply the number of ways there are to pick p objects among n different ones.  It's so commonly used when counting things that we need a special notation for this.  In handwriting or in print,  C(n,p)  in often written as a pair of parentheses enclosing an n above a p.  For example C(10,5) may be written as a 10 above a 5 between parentheses, without anything else within the parentheses:

C(10,5)   =   10C5   =   C 10
   =    ì
   =   252

How do we obtain this value of C(10,5)?  Well, let's first count how many ways you could pick the 5 elements in order:  You have 10 choices for the first element, but then there are only 9 unpicked elements left, so you only have 9 choices for the second element, 8 for the third, 7 for the fourth and 6 for the fifth.  All told, you have 10´9´8´7´6 = 30240 ways to choose an ordered sequence of 5 objects if you have 10 to choose from.  This is not quite the answer we want, because we're after the number of ways you could get a given set of 5 elements irrespective of the order in which you pick them...  Read on.

Let's count the number of ways to order 5 things.  It looks very much like what we did before:  You have 5 ways to pick the first, 4 ways to pick the second, 3 ways to pick the third, 2 ways to pick the fourth, and just 1 possibility left for the fifth and last.  All told, you have 1´2´3´4´5=120 ways to order 5 elements.

Each of your 30240 sequences of 5 objects is thus part of one (and only one) collection of 120 sequences in which the same objects appear, in some order.  The number of such collections is clearly 30240/120, or 252.  Therefore, there are 252 different ways to pick 5 unordered elements among 10: C(10,5)=252.

Make sure you understand the above example before going any further...

Instead of writing  1´2´3´4´...´(n-1)´n,  you may write n!  (it's called a factorial, but there's nothing to it; it's just shorthand).  The factorial notation makes it easy to express what we just told you about:  What's the number of ways you can order p elements?  Well, p!, of course...

The number of ways there are to pick an ordered sequence of p elements from n possible ones, is obtained with the same method used in the above example:

n (n-1) (n-2) ... (n-p+1)   =   n! / (n-p)!

Think about it:  The left-hand side has p factors and the right-hand side is a product of n factors divided by (n-p) factors so there are only p of them left...

In terms of factorials, C(n,p) equals the above divided by (p!).  Therefore:

C(n,p)   =   n! / [(n-p)!p!]

I hope I did not lose you.  If all these things are new to you, it's normal to get confused at first.  Study the above and everything will be clear, eventually...

The Binomial Theorem   (Khayyám's Theorem) :

The  choice numbers  C(n,p)  are also called  binomial coefficients  because they appear in the following formula for the  n-th  power of a  binomial.  The summation is understood to entail all integral values of  p  from  0  to infinity.  However, when  n  is an integer, the sum includes only n+1 terms, because  C(n,p)  is zero when  p  is greater than n.

(1+x)n   =   Sp  C(n,p) xp

To establish the validity of that formula, we simply remark that  xp  appears as many times in the expansion of the product of  n  factors on the left-hand side as there are ways to choose the term  x  (instead of the constant term 1)  in  p  of those factors.

This was first done by  Omar Khayyám (1048-1131).  To denote the unknown in his statements,  Khayyam used the Arabic term  transliterated  shay  or  xay  ("the thing").  This became the  "x"  of our current usage, examplified by the above formula !

Isaac Newton  (1643-1727)  remarked that Khayyam's formula remains valid even when the exponent  n  is not an integer, provided the right-hand side is duly interpreted as an  infinite sum  (a  power series  that converges when  |x| < 1 )  and  C(n,p)  is properly defined in the following way, which doesn't  require  n  to be an integer:  Isaac Newton 

C(n,p)   =     n (n-1) (n-2) ... (n-p+1)

 Blaise Pascal 

Pascal's Formula and Pascal's Triangle:

C(n+1,p+1)   =   C(n,p) + C(n,p+1)

This could be shown by using the above expression for C(n,p), but there's an easier approach...  Just notice that you can pick (p+1) objects among (n+1) in one of two ways:  Either you pick the last object or you don't!  There are C(n,p) ways to do the first thing and C(n,p+1) ways to do the second.  Halmos

With this formula, you can build a table of the  choice numbers  without any multiplications.  If you write the coefficients C(n,p) in a table where n is the number of the line (lowest on top) and p the number of the column (lowest to the left).  Each coefficient is the sum of the two numbers above it  (one is directly on top, the other is to the left of that one).

     0   1   2   3    4    5    6    7    8    9   10   11  12  13

 0:  1
 1:  1   1
 2:  1   2   1
 3:  1   3   3   1
 4:  1   4   6   4    1
 5:  1   5  10  10    5    1
 6:  1   6  15  20   15    6    1
 7:  1   7  21  35   35   21    7    1
 8:  1   8  28  56   70   56   28    8    1
 9:  1   9  36  84  126  126   84   36    9    1
10:  1  10  45 120  210  252  210  120   45   10    1
11:  1  11  55 165  330  462  462  330  165   55   11    1
12:  1  12  66 220  495  792  924  792  495  229   66   12   1
13:  1  13  78 286  715 1287 1716 1716 1287  715  286   78  13   1
14:  1  14  91 364 1001 2002 3003 3432 3003 2002 1001  364  91  14
15:  1  15 105 455 1365 3003 5005 6435 6435 5005 3003 1365 455 105
16:  1  16 120 560 1820 4368 8008 ...
17:  1  17 136 680 2380 6188 ...
18:  1  18 153 816 3060 8568 ...
19:  1  19 171 969 3876 ...

Note that C(n,n) is 1, so is C(n,0)  (there's only one way to pick no object at all).  More generally,  C(n,p) = C(n,n-p)  because there are as many ways to pick p objects as there are ways to discard the same number.

The above triangular table formed by the binomial coefficients is an ancient mathematical object whose origins have been traced to the second century BC  (in India).  It way well be much older than that.
It's associated with different mathematicians in different languages.  Local versions of the  Stigler's Law of Eponymy apply:  The Persians call it  Khayyám's triangle  although they know it was published a century before Khayyam, by  Al Karaji (953-1029).  Likewise, the Chinese call it the triangle of Yang Hui,  although Yang Hui himself attributed it to  Jia Xian (c.1010-c.1070).
In the West, it's usually called Pascal's triangle  but it's known to the Italians as  Tartaglia's triangle  (Niccolò Tartaglia published the first six rows in 1556).  Albert Girard (1595-1632) called it the  triangle of extraction  when he published it in 1629  (when Pascal was just 6 years old).  This emphasized the rôle of the coefficients in the algorithm of extraction of an n-th root, which was known to the medieval Chinese.  Pascal himself just called it  le triangle arithmétique.

Pascal's triangle  has many interesting properties.  Here are some of them:

Fibonacci Sequence :

It's not difficult to prove that sums of the successive  antidiagonals  form the celebrated  Fibonacci sequence :

1,   1,   1+1 = 2,   1+2 = 3,   1+3+1 = 5,   1+4+3 = 8,   1+5+6+1 = 13,   1+6+10+4 = 21,   1+7+15+10+1 = 34,   1+8+21+20+5 = 55     (A000045)

Gould's Sequence  (Dress Sequence) :

Count the number of  odd  entries in line number  n.  This gives you the sequence shown below, known as  Gould's sequence.  It was apparently introduced in 1961 by  Henry W. Gould (b. 1928) but Manfred R. Schroeder  (b. 1926)  has also called it the  Dress Sequence, after Andreas Dress (b. 1938).

1, 2, 2, 4, 2, 4, 4, 8, 2, 4, 4, 8, 4, 8... (A001316)

Every term in it is a power of  2.  The exponent in the power which corresponds to the  number n row  turns out to be the number of  "1"  bits in the binary expansion of  n  (A000120).  For example, the number  n = 1000000000  (one billion)  is expressed in binary as  111011100110101100101  which has  13 ones  in it.  Therefore, the number of odd entries in that distant row is:

 Waclaw Sierpinski 

2 13   =   8192

Pascal's Triangle  by  Casandra Monroe   (Numberphile, 2017-03-10).
Project Euler:  Problem 242   |   Sierpinski Triangle

(2013-11-26)   Multichoice Numbers
The number of ways to separate  n  objetcs into several collections.

The ordinary choice number  C(n,p)  can be seen as the number of ways to put n objects into either a red box or a blue box so that they contain respectively p and n-p objects.  Note that:

C(n,p) = C(n,n-p)     and     C(n,p) = 0  when either  p < 0  or  n < p.

Likewise, the number of ways to distribute  n  objects into three boxes so that the first two contain respectively  p  and  q  objects is denoted  C(n,p,q).  Such a distribution can be obtained by picking  p  objects among n for the first box and  q  objects among the remaining  n-p  for the second box  (the rest goes into the third box).  Clearly, the result doesn't depend on the order in which the boxes are filled and it can be expressed symmetrically:

C(n,p,q)  =  C(n,p) C(n-p,q)  =  C(n,q) C(n-q,p)  =  C(n,p+q) C(n-p-q,p)
C(n,p,q)   =   n! / [ (n-p-q)! p! q! ]

More generally, the number of ways  n  distinct objects can be distributed among  k  distinct boxes of fixed sizes  (n1 ... n)  is:

C(n, n1 ... nk-1 )   =   n! / [ n1! ... nk-1! nk! ]   where   n1 + n2 + ... + nk  =  n

Such a quantity is called a  multichoice number.

(2018-08-29)   How many ways to put  p  balls into  n  numbered bins?
Answer :   It's the  choice number   C(n+p-1 , p)   Let's prove that:

Consider putting markers into an array so that the leftmost square is marked and a total of  p  squares are left unmarked.  If the array has  n+p  squares,  there are  n  markers.  Because the position of one marker is imposed,  the number of ways to do so is:

C(n+p-1 , n-1)   =   C(n+p-1 , p)

Such arrays are clearly in one-to-one correspondence with the repartitions of  p  identical balls into  n  numbered bins:  We just associate bin number  k  with the  k-th  marker from the left.  The number of balls in the bin is the number of unmarked squares to the right of its assigned marker  (until either the next marker or the right end of the array,  whichever comes first).   QED

To put it in a nutshell:

  • There are  C(n,p)  ways to make  p  unordered choices among  n  possibilities,  if duplicates are disallowed.
  • There are  C(n+p-1 , p)  ways to make  p  unordered choices among  n  possibilities,  if duplicates are allowed.

This simple remark governs the statistical behavior of elementary particles,  according to  quantum logic.  Loosely speaking,  the above  bins  are  quantum states  and there are two possibilities:

  • There can be at most one  fermion  per bin.
  • There can be many  bosons  in each bin.

Quantum mechanics   |   Bose-Einstein statistics (1924)

(2013-04-23)   Outcomes with  p  Undistinguishable Dice
How many different outcomes are there from  p  dice with  n  faces.

Answer :   C ( n+p-1 , p )   as explained above.

Dice Theorem   (The many misconceptions in the "footnote on dice" are debunked here.)

calcuqueen (Linda of Hot Springs, SD. 2000-10-09)
A few days ago [2000-10-06], I asked how many 3-scoop sundaes I could make with 7 flavors of ice cream  (allowing repeated use of the same flavor).  I found the answer (84) by enumeration and someone posted the same result as 7 + (6)(7) + C(7,3).  I discovered this interesting pattern:
84   =   (1)(7) + (2)(6) + (3)(5) + (4)(4) + (5)(3) + (6)(2) + (7)(1)
(Taylor of Portland, OR. 2000-10-22)
How many triple scoop cones can be made with 31 flavors of ice cream?

For  n  different flavors, the answer is the following  choice number:

C(n+2,3)   =   n(n+1)(n+2)/6

That's  84  when  n = 7,  or  5456  when  n = 31.
That's a thinly disguised special case of the  previous problem.

To provide an alternate explanation in concrete terms,  let's consider that the flavors are listed in alphabetical order on the menu.  A  weird  waitress insists that you may not name any flavor more than once but are allowed to use the words FIRST and LAST (at most once each) to refer respectively to the first or the last flavor you picked explicitely.  Each type of sundae can be so named in a unique way.  This is equivalent to picking 3 different "flavors" among n+2  (including the two bogus "flavors" FIRST and LAST)  which can be done in C(n+2,3) ways.  QED

Now, about Linda's pattern  (which is indeed pretty).  It's fairly easy to prove its correctness using, for example, the formulas for sums of consecutive integers and squares of consecutive integers.  However, there's a better way, using straight combinatorics (which Linda may or may not have used to discover the pattern):

To name a sundae, you may first choose your "middle" flavor from the menu.  You are then allowed to pick one flavor below the middle one (including itself) and one flavor above it (including itself), for a total of 3 flavors, allowing any repetitions.

Thus, if your  middle choice  is  kth  on the list, you have  k (n+1-k)  possible choices;  k  choices for the  lower  flavor and  (n+1-k)  for the  upper  one.  The sum over all possible  middle choices  (from k=1 to k=n)  is the total number of possible sundaes.  That's also Linda's "pattern":

C(n+2, 3)   =    
k = 1
   k  (n+1-k)

Matthew LeBaron (2002-03-15; e-mail)
[...] Shirts come in 3 sizes (S, M, and L) and 5 colours (Red, Blue, Green, Black, and White). They are sold in packages containing 4 randomly selected shirts.  Order within a package is irrelevant and duplicates ares allowed.  How many types of packages are there?  What's the formula to determine this, given any set of numbers?

The number of different types of shirts to choose from is:

n   =   3 x 5   =   15

To specify a package of  p = 4  shirts,  you must select  p  of those  n  types,  allowing repetitions.  It's like putting  p  identical balls into  n  numbered bins.  The  number of ways to do so  is:

C(n+p-1 , p)   =   C(15+4-1 , 4)   =   C(18,4)   =   3060

That's the correct numerical answer to a silly question.  (What sane retailer would ever bundle together shirts of different sizes?)

As is often the case,  there are several ways to count the same thing and this may lead to interesting identities.  For example, we may build a duplicates-allowed choice of p items of n possible types by first choosing a certain number q+1 of different types  [for some q between 0 and p-1],  which can be done in C(n,q+1) ways.  Once this is done, we have C(p-1,q) possibilities to cast the chosen types on the p items. (HINT: You may do so by choosing at which q "intervals" in an ordered lineup of p items the item type is seen to "increase".)  All told:

C(n+p-1, p)   =    
q = 0
   C(n,q+1)  C(p-1,q)

(2004-02-14)   Leterrier's Problem
In general,  n  points of the plane define  C(n,2)  straight lines,  of which no two are parallel and no three meet outside of the  n  original points.  How many  new  points of intersection are there?

I am dedicating this to my high-school teacher, Mr. Leterrier  [also a  khôlleur  at Taupe Laplace]  who posed this nice question to my high-school class in 1972 or 1973  (time flies).  I still remember finding the answer through some awkward method, just like the rest of the class did.

Only after simplifying the result did it occur to me that there  ought to be  an  elegant solution.  Give it a fair shot before peeking! 

Sets with few ordinary lines (2013)  by  Terence Tao.

(2016-01-17)   Catalan Numbers
The result of many enumerations of interest to computer scientists.

1, 1, 2, 5, 14, 42, 132, 429, 1430, 4862, 16796, 58786...  (A000108).
Starting with  n = 0,  the  nth  Catalan number is   C(2n,n) / (n+1)

This quantity often occurs in combinatorics.  For example:

  • It's the number of different binary trees with  n  internal nodes.
  • It's the number of ways an n-sided convex polygon can be triangulated using only nonintersecting diagonals.

Eugène Catalan (1814-1894; X1833)

(J. F. of Los Angeles, CA. 2000-10-29)
Suppose 1.5 percent of the plastic spacers produced by a high speed mold injection machine are defective. For a random sample of 200 spacers, find the probability that:
A. None of the spacers are defective.
B. Three or more of the spacers are defective.

With p=0.015 and q=1-p=0.985, exactly k spacers are defective with probability C(200,k) pk´q200-k. In particular:

No spacer is defective (k=0) with probability q200, which is about 4.86683%.

Exactly one is defective (k=1) with probability 200´p´q199, or about 14.823%

Exactly 2 are defective with probability 19900´p2´q198, or about 22.46%.

Adding the above three results, we see that 2 or less spacers are defective with a probability of about 42.15%. Therefore, 3 or more are defective with a probability of about 57.85%.

Incidentally, the most likely number of defective spacers is k=3 (with a probability of about 22.574%), after which the probabilities decrease rapidly: 16.93% for k=4, 10.1% for k=5, 5% for k=6, 2.11% fo k=7, 0.76% for k=8, 0.25% for k=9, 0.073% for k=10, etc. In a sample of n spacers, the average number of bad spacers is n times the probability of a bad spacer. Here, that means 200´0.015, which is also 3.

goldda ( 2001-03-14)
How likely is it that 2 out of 5 children have the same birthday?

As usual, observe that "2 children have the same birthday" when 3, 4, or 5 do...  Let's first suppose that we are told that no one is born on February 29th:

If we assume each kid has one of 365 equiprobable birthdays, it's easier to compute the probability that they all have different birthdays:  The second kid has a birthday different from the first one with probability (364/365).  Knowing that the first two birthdays are different, the third birthday is different from these two with probability (363/365), etc.  All told, the probability of having 5 different birthdays is 364×363×362×361/3654. The probability that (at least) two kids share the same birthday is therefore 1 - 364×363×362×361/3654, or 481626601/17748900625, which is about 2.71% (0.0271355737).

Now, let's bring leap years and February 29 into the picture. We may even afford the luxury of using the full Gregorian calendar (century years are not leap years except when divisible by 400; 2000 was a leap year, 1900 was not).  This is appropriate only when we do not know at what time the family lived (for a 20th or 21st century family the Julian odds of exactly one leap year in 4 are more appropriate since 2000 was a leap year).  The Julian probability of having one's birthday on Feb. 29 is q = 1/1461 (1 leap year in 4 years).  The corresponding Gregorian probability is q = 97/146097 (97 leap years in 400 years).

Using either value for q, we may state that none of the 5 kids were born on February 29 with a probability (1-q)5 for which the above analysis applies, so two kids have an identical birthday other than Feb. 29 with a probability (481626601/17748900625)(1-q)5 To this, we should add the probability that at least 2 kids were born on Feb. 29, which is  1-(1-q)5.

All told, this gives a probability of 180043909073061/6656578833083301 or about 2.7047512782 % in the Julian case (limited to 20th/21st century), whereas the Gregorian case (from recent history to far into the future) corresponds to a probability of about 2.7050013288 %.  (That's exactly 1800420599383851642470257 chances in 66558954341646251642470257.) To summarize, the desired probability is about:

  • 2.705001 %   based on 97 leap years in 400 years.
  • 2.704751 %   assuming every 4th year is a leap year (1901-2099)
  • 2.713557 %   if we know no one's born on Feb. 29.

Does this result apply to 5 siblings?

No, it does not.  Heck, it doesn't even apply to a random group of real people, because maternity wards are notoriously busier at certain times of the year...  Various additional statistical biases apply to siblings.  One major observation is that twins are not so rare, especially among large families.  A minor observation is that the same woman cannot give birth in March and June of the same year...  For completeness, two people born on the same day of the same year may have the same mother, even if they're not twins.  (Guess how.)

By itself, the possibility of twins (in about 2% of live births) overshadows and invalidates the above result when applied to siblings.

As far as the laws of mathematics refer to reality, they are not certain,
 and as far as they are certain, they do not refer to reality.

Albert Einstein (1879-1955)

Jane (2007-08-22)    Covariance of two random variables  X  and  Y
cov ( X, Y )   =   E( [X-E(X)] [Y-E(Y)] )   =   E(XY) - E(X) E(Y)

cov(X,X)   is the  variance  of  X  (i.e., the square of its  standard deviation ).

By definition, the covariance of two random variables vanishes if and only if those two variables are  statistically independent  (i.e., the mathematical expectation of their product is the product of their respective mathematical expectations).

The notion can be illustrated by the example of two random variables X and Y obtained from three  independent  variables  (A, B, U)  and two  constants  (x,y).  Without loss of generality, we'll assume that the variance of U is 1.

X   =   A + x U
Y   =   B + y U

The covariance of X and Y is then seen to be  cov(X,Y)  =  xy. 

To compute that quickly, we introduce another independent variable U'  (distributed just like U)  and define  Y'=B+yU'  (thus, cov(X,Y') = 0).
  cov(X,Y)   =   cov(X,Y) - cov(X,Y')   =   xy E(U) - xy E(UU')   =   xy

deweydog (2001-03-24)   Variance
If you have a binomial distribution with n=200 and p=0.47, what would be the variance? How do you work this out?

If n independent random samples are either equal to 1 (with probability p) or to 0 (with probability 1-p), the sum of their values is a random variable which is said to have a binomial distribution; its average is np and its variance is np(1-p).  Here, this means 200´0.47´0.53, or 49.82.  (The standard deviation is the square root of this, roughly 7.0583.)

The explanation for this simple formula repays study.  Here it is:

To obtain a binomial distribution (both in theory and in practice), you may consider the sum Y of n independent random variables (X1, X2, ... Xn), each of which being equal to 1 with probability p and to 0 with probability q = 1-p. The sum is equal to k with probability C(n,k)(1-p)n-k pk . The "binomial" distribution derives its name from the binomial coefficients C(n,k) which appear in this expression.

The average ("expectation" or "expected value") E(Y) of a sum being the sum of the averages, we have simply E(Y)=np. (The expectation E is a linear function.) With n=200 and p=47%, the sum Y will have an expected value of E(Y)=94.

The variance V(Y) of Y is the expectation of the random variable (Y-E(Y)) 2, namely V(Y)=E([Y-E(Y)] 2=E(Y 2-2E(Y)Y+E(Y) 2)=E(Y 2)-E(Y) 2. (The last equality in this relation is known as Koenig's theorem. It is obtained simply by noticing that E(Y) is just a constant number. The expectation of a constant number is itself and the expectation of a random variable multiplied by any constant number is simply that constant number multiplied by the expectation of the random variable.)

For two independent random variables A and B, the relation E(AB)=E(A)E(B) holds (which is not generally the case for dependent variables). You may consider the variance of the sum A+B: V(A+B)=E([A+B] 2-[E(A+B)] 2. V(A+B) may thus be expressed as E(A 2+B 2+2AB)-[E(A)+E(B)] 2. Since we have E(AB)=E(A)E(B) for statistically independent variables, this boils down to E(A 2)+E(B 2)-E(A) 2-E(B) 2, which you may recognize as equal to V(A)+V(B). In other words, if A and B are independent, V(A+B)=V(A)+V(B).

Therefore, the variance of the sum Y is n times the variance of each independent term X. What is the variance of X? Well, as X is either 0 or 1, X 2=X and E(X 2)=E(X)=p so that V(X)=E(X 2)-E(X) 2=p-p 2=pq.
All told, V(Y) = npq , as advertised.

(Seetharaman, United Kingdom. 2000-10-31)   Estimating Variance
In standard deviation calculations, why are we using N-1 instead of N?
There are some books where I find N being used instead of N-1.

Both formulas are correct but they apply to different situations. Basically, you use N when you know the theoretical average of the sample (for example, you may know it's zero because of some symmetry in the phenomenon), whereas you use N-1 when you are working with an estimate of the average, obtained by averaging the sample at hand.

This latter use of (N-1) comes from the fact that you want what's called an "unbiased" estimator of the standard deviation, namely you want the calculations to yield a result which does not have a systematic built-in error...

Consider n samples X1, X2, ... Xn obtained from an unknown random variable X.  The expected value of the sum is the sum of the expected values. Therefore, if you want to estimate the mean of X, your best shot is to divide by n the sum of the samples. Call this the "sample average".  It's an "unbiased" estimate of the mean in the sense that the expected value of the "sample average" equals the true mean of the random variable X.  So far so good.

Now, if you want to estimate the standard deviation but do not know the true mean of X,  you'll have to use the best estimate of the mean at your disposal, namely the sample average.  The problem is that this sample average is obtained from the sample itself and is therefore not independent of each individual sample.  The quadratic deviation of X1 from the sample average is:


The expected value of that quantity may be computed from the fact that all the samples are independent and have the same expected value  E(X).  We obtain:

(1-1/n)2 E(X2) + (n-1)/n2 E(X2) - 2(1-1/n)(n-1)/n E(X)2 + (n-1)(n-2)/n2 E(X)2

This boils down to   (n-1)/n [E(X2)-E(X)2].   Now, the quantity E(X2)-E(X)2 equals the square of the standard deviation s.  In other words, the square of the deviation of X1 from the sample average is expected to be (n-1)/ns 2

What's true of  X1  also applies to  X2,...,Xn  and, therefore, the sum of the  n  samples has an expected value of  (n-1)s 2.

That's why you have to divide the sum by (n-1) instead of n when you're working with the sample average because the true mean is unknown.

When the true mean is used, there are no such complications from interdependencies of the quantities involved and the divisor n should be used.

In other words, we may estimate the standard deviation s of a random variable X from a set of n samples X1, ..., Xn using either of the following unbiased formulas:

  • s 2 = [ X12 + ... + Xn2 - n ´ m2 ] / n
  • s 2 = [ X12 + ... + Xn2 - n ´ A2 ] / (n-1)

In this, m is the true mean of X, whereas A = (X1+...+Xn)/n  is an estimate of m based on the sample at hand.

The relevant nontrivial factor   1-1/n   is known as  Bessel's correction.

(2007-08-28)   Markov's Inequality
If  Y  is a non-negative random variable, then   P( Y ≥ a)  ≤  E(Y) / a

Proof :   If Y ≥ a > 0,  then 1 ≤ Y/a.  So, the function which equals 1 when Y ≥ a and 0 elsewhere has a lesser mathematical expectation than  Y/aQED

Markov's inequality   |   Andrey Markov (1856-1922) inherited it from Chebyshev (his teacher).

(2007-08-28)   Bienaymé-Chebyshev Inequality:  P( |X-m| ≥ ks) ≤ 1/k2
For  any  random variable  X  of mean  m  and standard deviation  s.

Proof :   Apply Markov's inequality to the random variable  Y = (X-m)2

P( |X-m| ≥ ks)   =   P( Y ≥ k2 s2 )   ≤   E(Y) / k2 s2   =   1 / k2     QED

The French mathematician Irénée-Jules Bienaymé (1796-1878; X1814) was a personal friend of Pafnuty Chebyshev (1821-1894) whose work he translated from Russian into French.  However, the inequality which now bears both names was actually first published by Bienaymé in 1853.  Chebychev would use it only 14 years later, in a generalized proof of the Law of large numbers.

MIMI192002 (2002-02-24)     Inclusion-Exclusion Enumeration
What is the probability P(AUBUC) of the union of 3 events A,B,C?

P(AUBUC)   =   P(A) + P(B) + P(C) - P(AÇB) - P(BÇC) - P(CÇA) + P(AÇBÇC)

You may prove this using the relation P(AUE) = P(A) + P(E) - P(AÇE), with E=BUC, namely: P(AUBUC) = P(A) + P(BUC) - P((AÇB)U(AÇC)). The last two terms may be expanded using the formula for the probalitity of a union again:
P(AUBUC) = P(A) + P(B) + P(C) - P(BÇC) - P(AÇB) - P(AÇC) + P((AÇB)Ç(AÇC)). The last term of this expression is clearly P(AÇBÇC).

This identity is a special case of a relation which is worth committing to memory: The probability of a union is equal to:

  • the sum of the probabilities of the individual events,
  • MINUS the sum of the probabilities of the intersections of 2 events,
  • PLUS the sum of probabilities of the intersections of 3 events,
  • MINUS the sum of probabilities of the intersections of 4 events,
  • PLUS etc.

This is easily proved by induction; the formula for n+1 events is derived from the formula for n events just as we went from 2 to 3 events in the above.

This type of computation is often called an inclusion-exclusion enumeration.  It holds either for the probability of events, or for the number of elements in sets.  Here are a few great examples:

1.   Probability that a permutation has at least one fixed point :

Among equiprobable permutations of  n  elements, the probability that  p  given elements are fixed points is   (n-p)! / n!   (since there are  (n-p)!  ways to choose a permutation of the elements other than those fixed points).  Thus, the  sum of the probabilities  of all such events is simply  1/p!  (namely, the above probability multiplied into the  C(n,p)  ways to choose  p  elements).

Clearly, there's  at least one fixed point  when point  i  is fixed for  some  i  between  1  and  n.  The probability of the union of those  n  nonexclusive  events is thus readily obtained by  inclusion-exclusion enumeration  using the remark presented in the previous paragrah:  The probability that there is  at least one fixed point   in a random permutation of  n  elements is thus:

1/1!  -  1/2!  +  1/3!  -  1/4!  +   ...   -  (-1)n / n!

Except for very small values of  n,  that expression is extremely close to a number which all electrical engineer memorize as 63.2% for daily use:

1-1/e   =   0.632120558828557678404476229838539...

This is to say that the probability that there are  no fixed points  in a permutation of many elements is very nearly equal to:

1 / e   =   0.367879441171442321595523770161460867445811131...

2.   Probability that a random function has no fixed points :

The reader may want to use the same reasoning to obtain the probability that a function  (not necessarily bijective)  from  {1...n}  to  {1...n}  has no fixed point  (using the fact that  p  given elements are fixed points with probability  1/np ).  The result is a binomial expansion which turns out to be equal to:

( 1 - 1/n ) n

Of course, that expression is more readily established by observing that a function without fixed points is merely obtained by choosing an image among  (n-1)  acceptable points for each of the  n  elements.

This simple argument makes  inclusion-exclusion enumeration  look awkward.  However, there was no counterpart to this shortcut in the previous case involving permutations  (i.e.,  bijective functions).

The limit of that expression as  n  gets large happens to be also equal to  1 / e  but the convergence is  much  slower than in the case of permutations:

( 1 - 1/n ) n   =    e n Log(1-1/n)   =   1/e   [ 1  -  1/2n  -  5/24n2  -  ...  ]

3.   Project Euler  Problem 239   (2009-04-03) :

What is the number of permutations of the integers 1 to 100 which leave exactly  3  of the  25  primes unchanged?  Answer :   [by inclusion-exclusion]

C(25,3)   22   (-1)n C(22,n) (97-n)!    =    1.761863476208905... 10155
n = 0

The probability that a random permutation has the above property is equal to that 156-digit integer divided by  100!  (namely:  0.00188785484103034...).

4.   Project Euler  Problem 193   (2008-05-10) :

What is the number of squarefree integers not exceeding  N ?

Let's exclude from the total count  (N)  all the multiples of the  squares of primes,  using a straightforward  inclusion-exclusion  approach:

For  q2  no greater than  N,   we're dealing with  floor (N/q)  multiples of  q2.  That number is either  excluded  (minus sign)  when  q  is a squarefree integer with an odd number of prime factors or included  (plus sign)  if  q  is a squarefree integer with an even number of prime factors.  If  q  is the multiple of a square, it's simply ignored, because the multiples of  q  are already dealt with in the above scheme  (as multiples of squarefree numbers).  All told, the desired result may be very neatly expressed using the  moebius function  m  which is precisely equal to  -1  in the first case and  +1  in the second  (it's  0  for a multiple of a square):

Ö N     m(q)   floor ( N / q)
q = 1

For  N = 2 50,  the value of this expression  (namely, 684465067343069)  can be computed in minutes  (assuming  m  to be decently implemented).

The value of  m  for millions of integers can be generated extremely fast as a large array of two-bit entries  (4 possible codes stand for the 3 legitimate values -1, 0 and +1 and a "not yet computed" place-holder, if needed)  This involves a slight modification of the  Sieve of Eratosthenes  which can be carried out without any multiplications or divisions, as described elsewhere on this site.

A computation based directly on the following naive expression  would be  hopeless  (it involves  1125899906842624 terms, instead of  33554432).

N     | m(q) |
q = 1

The numbers of squarefree integers not exceeding  10n  (for n = 0, 1, 2, ... 17)  appear on the vertical list at left  (A071172).  This reveals the string of digits  (A059956)  given on the last line which is just the decimal expansion of the ultimate density of squarefree integers among natural integers  (about 60.79% of integers are squarefree)  namely:

6 / p2   =   0.607927101854026628663...
=   (1 - 1/22 ) (1 - 1/32 ) (1 - 1/52 ) ... (1 - 1/p2 ) ...


Beautiful math (2016)  &  UCLA Math Distinguished Lecture (2015-05-21)  by  Manjul Bhargava (1974-).

5.   Cubefree numbers not exceeding N :

The same approach would yield  (even more efficiently)  the number of  cubefree  integers not exceeding N, which is given by the formula:

N1/3     m(q)   floor ( N / q)
q = 1
In the above,  N1/3  actually stands for the largest integer whose cube doesn't exceed  N.  To obtain this precise value with ordinary floating-point arithmetic requires some care.  A "safe" expression would be:

ë N 1/3 û   =   floor ( (N+0.5)1/3 )

In particular, when  N  is equal to the successive powers of  10  we obtain the following sequence  (A160112)  in whose digits appear the decimal expansion of the density of cubefree integers  (namely, the reciprocal of  Apéry's constant).

1, 9, 85, 833, 8319, 83190, 831910, 8319081, 83190727, 831907372, 8319073719, 83190737244, 831907372522, 8319073725828, 83190737258105, 831907372580692, 8319073725807178, 83190737258070643, 831907372580707771, 8319073725807074649, 83190737258070746354, 831907372580707468393, 8319073725807074687961, 83190737258070746868817, 831907372580707468685592, 8319073725807074686838076, 83190737258070746868310774, 831907372580707468683127869, 8319073725807074686831257909, 83190737258070746868312631845 ...

1/z(3)   =   0.831907372580707468683126278821530734417...   A088453

Squarefree Numbers:  A071172, A053462, A143658   |   Cubefree Numbers:  A160112, A160113
Counting Square-Free Numbers  by  Jakub Pawlewicz   (2011-07-25).

WiteoutKing (Lowell, MA. 2002-07-17)   [ Cf. abridged answer. ]
What are the odds in favor of being dealt a given poker hand?

There are C(52,5) = 2598960 different poker hands and each of them is dealt with the same probability.  [The choice number C(52,5) is pronounced "52 choose 5" and is equal to the number of ways 5 objects can be picked among 52 when the order is irrelevant, as explained above.]

Therefore, it's only a matter of counting how many such hands belong to a given type to determine the probability of that type (namely, the ratio of that number to 2598960, the total number of possible hands).  When the probability of an event is the fraction P = x / (x + y) , its so-called odds are said to be either x to y in favor or y to x against.  We pay tribute to this popular way of expressing probabilities in the list below (which assumes some familiarity with poker hands).  Note that there are normally 10 different "heights" for a straight and that the ace (A) belongs to the lowest (A,2,3,4,5) and the highest (10,J,Q,K,A), which is traditionally called a Royal Flush if all cards belong to the same suit.

  • Royal Flush: C(4,1)C(1,1) = 4;  P = 1/649740 ("1 to 649739 in favor") 
  • Straight Flush: C(4,1)C(9,1) = 36;  P = 3/216580 ("3 to 216577") 
  • 4 of a Kind:  C(13,1)C(48,1) = 624;  P = 1/4165 ("1 to 4164") 
  • Full House: C(13,1)C(4,3)C(12,1)C(4,2) = 3744;  P = 6/4165 ("6 to 4159") 
  • Flush: C(4,1) [C(13,5) - 10] = 5108;  P = 1277/649740 ("1277 to 648463") 
  • Straight: C(10,1) (45-4) = 10200;  P = 5/1274 ("5 to 1269")
  • 3 of a Kind: C(13,1)C(4,3)C(12,2) 42 = 54912;  P = 88/4165 ("88 to 4077")
  • Two Pairs: C(13,2)C(4,2)2C(44,1) = 123552;  P = 198/4165 ("198 to 3967")
  • Pair: C(13,1)C(4,2)C(12,3) 43 = 1098240;  P = 352/833 ("352 to 481")
  • High Card: (C(13,5)-10) (45-4) = 1302540;  P = 1277/2548 ("1277 to 1271")

A word of explanation for the last count.  "High Card" means that the hand is "none of the above"  (two such hands would be compared highest card first to decide who wins).  These hands are uniquely determined from two independent choices:  First, pick 5 distinct values in one of the  [C(13,5)-10]  ways which aren't straights, then assign suits in one of the  [45-4]  ways that aren't flushes.

The total of all of the above is 2598960 = C(52,5), because every hand is of  one and only one of the 10 types listed.  You may want to check this...

Should your own local rules disallow the 10th straight sequence (A,2,3,4,5), the above counts for straights and/or flushes should be changed (and the "High Card" count should be modified as well), replacing 4, 36, 5108, 10200 and 1302540 respectively by 4, 32, 5112, 9180 and 1303560 = (C(13,5)-9)(45-4).

Don Anglen (2004-01-26)     7-Card Stud Poker
I'm told there are 41584 seven-card hands [out of C(52,7)] containing one of the 40 five-card straight flushes.  Why not 40 C(47,2) = 43240 ?

Let's first count how many ways there are to get a royal flush (A,K,Q,J,10 in the same suit) when you are dealt q cards.  Among the C(52,q) hands of q cards, the following number contain a royal flush  (recall that C(n,p) is 0 for a negative p):

C(4,1)C(52-5,q-5) - C(4,2)C(52-10,q-10)
+ C(4,3)C(52-15,q-15) - C(4,4)C(52-20,q-20)

The first term corresponds to a choice of one of the 4 possible royal flushes followed by q-5 cards among the 47 remaining cards.  This would count more than once any hand with more than one royal flush in it, so a negative adjustment must be made...  What you must do is a proper "inclusion-exclusion" enumeration (as presented above) which is where the 3 additional terms come from.

The same thing goes for straight flushes (let's consider a royal flush to be a particular case of a straight flush) except the alternating terms are more complex, and more numerous  [as many as 20 straight flushes could coexist in a hand of q = 26 cards].  Clearly, the "naive" 40 C(47,2) count  (for q = 7)  tallies more than once any hand containing more than one straight flush...

Below is a correct enumeration.  The n-th square bracket corresponds to the total number of hands containing n straight flushes.  The bracket contains up to 4n-3 terms, each corresponding to a given total number of cards used by the n straight flushes under consideration (this number can be anything from 4+n to 5n).  The coefficients in front of the binomial numbers add up to C(40,n):

  [ 40 C(52-5,q-5) ]
- [ 36 C(52-6,q-6) + 32 C(52-7,q-7) + 28 C(52-8,q-8) +
    24 C(52-9,q-9) + 660 C(52-10,q-10) ]
+ [                  32 C(52-7,q-7) + 56 C(52-8,q-8) + ... ]
- [                                   28 C(52-8,q-8) + ... ]
+  ...

For example, the second of the above brackets corresponds to two given ("featured") simultaneous straight flushes (there may be more, we don't care):  Among the C(40,2) = 780 such featured pairs, 36 use 6 cards, 32 use 7 cards, 28 use 8 cards, 24 use 9 cards, and 660 pairs are disjoint and use 10 cards.  Many of the terms cancel out (we're missing a shortcut here) so the formula can be greatly simplified, but there's no need for sophistication in the case where q = 7, since the terms already given explicitely include all the nonzero ones.  For q = 7, the thing boils down to the advertised result:

40 C(47,2)  -  36 C(46,1)   =   41584

Don Anglen (2004-01-26)   26-Card Stud Poker   &   q-Card Poker
What's the probability of getting a straight [or royal] flush with 26 cards?

The enumeration presented in the previous article is only easy for small hands or very large ones:  For example, in a "hand" of  q = 45  cards there's always at least one straight flush.  Only one hand of 44 cards is without a straight flush.

However, no such shortcut applies to the case  q = 26.  We'll use a simple computer program to do the actual counting fairly efficiently, in two steps:

Step 1:

First, let's consider the C(13,q) hands consisting of q cards of the same suit.  Let's call N(q) the number of these which contain at least one straight.  There are only 8192 different such hands for q between 0 and 13 (that's 2 to the power of 13) so it's easy to tabulate N(q) in a fraction of a second with a computer.

To do so, we assign each hand a pattern of 13 bits (corresponding to the integers from 0 to 8191) with the most significant bit set if we hold an Ace, the next one for a King, and so on, down to the least significant bit for a deuce.  The number 31 (i.e, 5 consecutive bits set to 1) multiplied by any power of two, from 1 to 256, gives a so-called mask corresponding to one of the 9 regular straights, while the mask for the tenth straight (A,2,3,4,5) is 4096+15 = 4111.  The tiny table below may thus be built by counting the number q of nonzero bits in each of the 8192 patterns whose "bitwise and" with at least one of the 10 acceptable masks is equal to that particular mask.  Here's a piece of UBASIC code to do just that:

   dim N(13)
   for I=0 to 8191
      if 4111=bitand(4111,I) then *TALLY
      while M<8192
         if M=bitand(M,I) then *TALLY
      wend: goto *SKIP
      *TALLY: N(Q)=N(Q)+1
   *SKIP: next I
Number of combinations of q cards from the same suit
(of 13 cards)  which contain at least one straight flush:
q< 55 6789 10111213 > 13
 N(q) 01071 217371384 234771310

Step 2 (using several methods):

Now, for the entire pack of 52 cards, we may consider the 4 nonexclusive possibilities of having straight clubs, straight hearts, straight diamonds, or straight spades...  Thus, a straightforward inclusion-exclusion enumeration gives the desired result as the sum of the following expressions  [of four terms]  over all the nonnegative indices q1 q2 q3 q4 whose sum is equal to q:

C(4,1) N(q1) C(13,q2) C(13,q3) C(13,q4)
- C(4,2) N(q1) N(q2) C(13,q3) C(13,q4)
+ C(4,3) N(q1) N(q2) N(q3) C(13,q4)
- C(4,4) N(q1) N(q2) N(q3) N(q4)

Such terms are zero when an index is more than 13 or when the function N appears with an argument below 5, so there are relatively few nonzero terms to add (or subtract):  For  q = 26,  the 4 above types of terms are nonzero in 1209, 709, 334 and 84 cases, respectively.  A grand total of 2336 nonzero terms... A priori, C(q+3,3) ordered sets of indices of sum q, and 4 terms to each set could have meant 14616 terms for q = 26.  All told,  255774304012724 hands contain a straight flush out of  495918532948104 possible hands of 26 cards.  Thus, the probability to have a straight [or royal] flush among 26 cards is:

9134796571883 / 17711376176718   =   0.5157587124082935...

Number of hands of q cards with at least one straight flush  (A143314)
and percentage of that to the total number of combinations   C(52,q) :
q=                   (%)
  4               0   0
  5              40   0.001539077 
  6            1844   0.009057633
  7           41584   0.031082810
  8          611340   0.081237077 
  9         6588116   0.179069883
 10        55482100   0.350708060
 11       380126920   0.629310354
 12      2177910310   1.055294393
 13     10644616240   1.676281723
 14     45049914588   2.546680140
 15    167011924492   3.726795587
 16    547315800984   5.281342552
 17   1597026077496   7.277207845
 18   4173458163098   9.780325218
 19   9813490226056  12.851547041
 20  20841357619302  16.541465273
 21  40096048882028  20.884249209
 22  70043238025686  25.890741583
 23 111299858400784  31.541291220
 24 161084365419139  37.779089838
q=                   (%)
 25 212534592698472  44.505092157 
 26 255774304012724  51.575871241
 27 280828262657348  58.805898612
 28 281310102228657  65.975612201
 29 257052287198104  72.846103901
 30 214209339496536  79.180215104
 31 162748148903084  84.768275047
 32 112708315365631  89.454857917
 33  71138289660672  93.161256080
 34  40921413681848  95.897626995
 35  21453967174040  97.759817712
 36  10250154666092  98.909218234
 37   4460732889232  99.539237677
 38   1766089532348  99.837373263
 39    634724725868  99.954515344
 40    206360120020  99.990654664
 41     60402956200  99.998720874
 42     15820006672  99.999889077
 43      3679075192  99.999994346
 44       752538149  99.999999867
 45       133784560 100

The above summation can be simplified into the following expression, which saves multiplications.  It involves only 508 nonzero terms for q = 26,  and never more than 1195 nonzero terms for any q  (this maximum is reached for q = 35):

åi N(i) [ 4 C(39,q-i) - åj N(j) [ 6 C(26,q-i-j) - åk N(k) [ 4 C(13,q-i-j-k) - N(q-i-j-k) ] ] ]

We could also use the elegant approach presented below and introduce the following polynomial S(x) which encodes the result of the above "first step":

10 x5 + 71 x6 + 217 x7 + 371 x8 + 384 x9 + 234 x10 + 77 x11 + 13 x12 + x13
=   x5 (1+x) (10 + 61 x + 156 x2 + 215 x3 + 169 x4 + 65 x5 + 12 x6 + x7 )

The number of combinations of q cards containing at least one straight [or royal] flush is then equal to the coefficient of  xq  in the following expression:

4 S(x) (1+x)39   -   6 S(x)2 (1+x)26   +   4 S(x)3 (1+x)13   -   S(x)4

This expression may be rewritten in a much more compact form:

(1+x)52   -   [ (1+x)13 - S(x) ] 4

The hands  without  a straight flush are thus enumerated by

[ (1+x)13 - S(x) ] 4

This is easy to prove:  The inside of the bracket counts the combinations of cards of the same suit without a straight.  A combination from the whole pack without a straight flush is just a juxtaposition of four such combinations; one in each suit.

Similarly, the formula given above for the number of combinations of q cards containing at least one  royal  flush is equal to the coefficient of   xq   in:

(1+x)52   -   [ (1+x)13 - x5 (1 + x)8 ] 4

So, the following difference counts hands with straight flushes but no royal ones:

[ (1+x)13 - x5 (1 + x)8 ] 4   -   [ (1+x)13 - S(x) ] 4

This is equal to  36 x5 + 1656 x6 + 37260 x7 + 546480 x8 + 5874656 x9 + 49346350 x10 + 337176880x11 +   ...   + 490000 x47+ 25000 x47 + 625 x48.
These coefficients appear on the second line of the table in the next article...

(2004-02-06, updated 2006-07-28)
What's the probability of a poker hand appearing highest among q cards?

Among q cards, there are 13  nonexclusive  ways to have at least one 4-of-a-kind hand  (4 aces, 4 kings ... 4 deuces).  Therefore  (cf. previous article)  the number of such q-cards combinations is the coefficient of  xq  in the polynomial:

(1+x)52 - [ (1+x)4 - x4 ] 13

However, when  q  is 8 or more, there's a possibility of having  both  a 4-of-a-kind and a straight flush, in which case the existence of the latter overshadows the former  (according to the careful wording of the above question).

For example, the coefficient of  x8  in the above (i.e., 2529462) includes 200 straight (or royal) flushes:  Each such hand is obtained from any of the 40 possible straight flushes by singling out one of its 5 cards [and including all other cards of its kind].  So, only 2529262 eight-card hands are counted as 4-of-a-kind.

Some of the data tabulated below may thus be surprising at first.  For example, with 8 cards or more, a full house is more likely than just three-of-a-kind...

The probability is the ratio of the tabulated number to the total on the last line.
q-Card Studq = 5q = 6q = 7 q = 8q = 9
Royal Flush 4188432464860713460
Straight Flush 361656372605464805874656
4 of a Kind624 14664224848 252926222247616
Full House3744 1659843473184 45652128423908824
Flush5108 2057924047644 50850320453008864
Straight10200 3616206180020 67072620509071920
3 of a Kind54912 7321606461620 38493000151728780
Two Pairs123552 253281631433400 2577609001442570040
Pair1098240 973074058627800 236092500600163200
High Card1302540 661290023294460 5347608069788040
TOTAL2598960 20358520133784560 7525381503679075400

Some of the above entries courtesy of  Bill Butler   |   The Wizard of Odds

With 11 cards, what's really difficult is to avoid getting at least a pair,  for you must hold  AKQJ9876432  (the only set of 11 different values which does not contain a straight)  without holding more than 4 cards of any suit.  There are as many ways to do this as there are arrangements of 11 suits where no suit appears more than 4 times.  As explained elsewhere on this page, this amounts to:

The coefficient of   x11 / 11!     in     ( 1  +  x2 / 2!  +  x3 / 3!  +  x4 / 4! ) 4

This coefficient is 11! times 25/432, which is 2310000  (whereas there are 411, or 4194304 unrestricted arrangements).  All told, there are thus 2310000 combinations of 11 cards that do not include a pair or better.  Out of a total of C(52,11) = 60403728840 equiprobable combinations, this is a probability of only 0.000038242672...  With 12 cards, this probability would be zero, since you would always hold either a straight or several cards of the same kind.

(Lizzie of United Kingdom. 2000-10-29)
There are 3632428800 different possible arrangements of the letters in the word "CONSTANTINOPLE".  How can I show that there are 457228800 arrangements of the letters so that no two vowels are next to each other?

Let's first count the unrestricted arrangement of the letters in CONSTANTINOPLE. There are 14 letters (3 N's, 2 O's, 2 T's, and a single copy of each of the 7 letters C,S,A,I,P,L,E).  If all these letters were different (say the N's are colored red white or blue, whereas both the O's and the T's are either red or blue), you would have 14! different arrangements.  To each actual arrangement of the (uncolored) letters corresponds exactly 3!2!2! arrangements of the colored letters.  Therefore, there are 14!/(3!2!2!) or 3632428800 possible arrangements of the (uncolored) letter in CONSTANTINOPLE.  So far so good.

What happens when you don't allow adjacent vowels?  You have 5 vowels and 9 consonants.  In how many ways can you put P vowels in 2P-1+K positions?

Well, if K is negative you can't do it. If K is zero, you've got only one choice (think about it). In general, you have C(K+P,P) choices, because that's the number of ways you could place the vowels in P+K positions if you did not have any restriction.  With the restriction, each vowel is in effect occupying two spaces (itself and the position to its right) except for the last one which is only occupying one space.  So, you may as well mentally ignore the P-1 "dead spaces" and see the exact correspondence between the two situations.

Here, you have P=5 and K=5, which means C(10,5) = 252 ways of choosing the positions of the vowels.  Once this is done, you have 9!/(3!2!) choices for placing the consonants (since there are 3 identical N's and 2 identical T's) and 5!/2! choices for the vowels (there are two O's). All told, you have 252 ´ 9! ´ 5! / 24  =  457228800  possibilities, as advertised.

Incidentally, the probability that no two vowels are adjacent in an arrangement of P vowels among N letters (N = 2P-1+K) depends only on N and P, because it's the same as the probability of having no adjacent marks when P squares are randomly marked in a row of N squares.  (Putting equiprobable arrangements of vowels in the marked squares and equiprobable arrangements of consonants in the unmarked ones won't change the odds.)  This probability is:

C(N-P+1,P) / C(N,P)

With P=5 and N=14, this ratio is simply 252/2002, or 18/143 (about 12.59%) which is an easier way to find the ratio of the above pair of large numbers...

(Lizzie of United Kingdom. 2000-10-31)     Generating Functions
From the 9 letters of the word "POSSESSES", show that there are...
a)   12 possible selections of 4 letters,  and
b)   115 different arrangements of 4 letters in a straight line.

The 9 letters include:  5 S's, 2 E's, 1 P, 1 O.

Let's use this question to present a very interesting general approach...

To make a selection, you have to chose a certain number of S's from 0 to the maximum allowed, which is 5 (the fact that you will eventually use at most 4 S's is irrelevant at this point), then you choose the number of E's from 0 to 2, and the numbers of P's and O's each from 0 to 1.

The four numbers you pick must add up to  4.  Well, the coefficient of x4 in the following expression is obtained with exactly the same breakdown, so both quantities are equal  (think about it)!

(1 + x + x2 + x3 + x4 + x5) (1 + x + x2) (1+x) (1+x)

In the expansion of this polynomial, the coefficient of  xp  is thus the number of selections of  p  letters taken from  POSSESSES:

1 + 4x + 8x2 + 11x3 + 12x4 + 12x5 + 11x6 + 8x7 + 4x8 + x9

In particular, the coefficient of  x4  is  12,  the answer to question (a).

Now, let's try to count arrangements with the same polynomial idea. Well, if all the letters were different, you could rearrange each of the above basic selections in 4! different ways. They are not different, though, and that means that for each set of k identical letters in a basic selection, you've overestimated the true count of arrangements by a factor of k!  All told, you'll see that the number of possible arrangements is 4! times the coefficient of x4 in the following polynomial (it's more convenient to describe this as "the coefficient of x4/4! "):

(1 + x + x2/2! + x3/3! + x4/4! + x5/5!) (1 + x + x2/2!) (1+x) (1+x)

Now, expand this polynomial and you get:

1 + 4x + 14x2/2! + 43x3/3! + 115x4/4! + 266x5/5! + 543x6/6! + 987x7/7! + 1512x8/8! + 1512x9/9!

In particular, the coefficient of x4/4! is 115 and that's the answer to your second question.

Incidentally, notice that, if you had an unlimited supply of the 4 letters, the above polynomial would become a series equal to the product of 4 identical series, each equal to exp(x). Therefore, the whole thing would simply be the series for exp(4x), namely:

1 + (4x) + (4x)2/2! + ... + (4x)n/n! + ...

This is one (devious) way to prove that there are 4n arrangements of n letters from an "alphabet" of only 4. (4 choices for the 1st position, 4 for the 2nd, 4 for the 3rd, 4 for the 4th, etc.)

I encourage you to do the same thing to count basic selections (not arrangements) of n letters of 4 possible differents kinds (whose supply is unlimited). What, then, is the coefficient of xn in 1/(1-x)4 ?   [Answer: C(n+3,3)]

Anonymous Query via Google (2004-11-10) :     How many positive integers less than 1000000 have the sum of their digits equal to 19?

The answer is 30492.  The approach introduced above shows that this is the coefficient of x19 in the expansion of  ( 1 + x + x2 + ... + x9 ) 6.

In UBASIC:   print coeff(((_X^10-1)\(_X-1))^6,19)

(2004-07-17)   Polynacci Numbers
How many sequences of n bits are there without p consecutives zeroes?

Equivalently, you could ask about the number of ways to flip a coin n times without ever getting p consecutive tails.  Same problem.

Let's call a bit sequence [of zeroes and ones] without p consecutive zeroes "satisfactory" and let F(n) be the number of such satisfactory strings of length n.

Since all sequences of length less than p are satisfactory, we have  F(n) = 2n,  for n<p.  On the other hand, if n is at least equal to p, a satisfactory sequence starts with exactly k zeroes (k = 0, 1, ... p-1), followed by a "one" bit and then any satisfactory sequence of length (n-k-1).  Therefore, the following relation holds, which defines F(n) recursively:

F(n)   =   F(n-1) + F(n-2) + ... + F(n-p)

In particular, for p=2, we're counting the number of flip sequences where tails never occurs twice in a row, and obtain as F(n) the Fibonacci number of rank (n+2).  For p=3, it's the so-called "Tribonacci" number of rank (n+3).  For p=4, we have "Tetranacci" numbers, for p=5 "Pentanacci" numbers, etc.  Note that, for p=1, F(n) = 1, because there's only one string of n bits without zeroes...

Number of strings of n bits without p consecutive zeroes :  (A048887)
n =01234 56789101112 Sloane's
p = 1 111111 1111111 A000012
p = 2 12 35813213455 89144233377 A000045
p = 3 124 71324 44811492745049271705 A000073
p = 4 124 81529 5610820840177314902872 A000078
p = 5 124 81631 6112023646491217933525 A001591
p = 6 124 81632 6312524849297619363840 A001592
p = 7 124 81632 64 127253504100420003984 A066178
p = 8 124 81632 64128 255509101620284048 A079262
p = 9 124 81632 64128256 511102120404076 A104144
¥ 1248163264128 256512102420484096 A000079

As there are 1024 equiprobable ways to flip a fair coin 10 times, the probability you'll do so without ever getting tails twice in a row is  144/1024 (about 14 %).

Michael J. Wales (2010-09-24; e-mail)   Endless Coin Flipping
If I  win  when I flip more  consecutive  heads than I've ever flipped tails, what's my probability of winning?

Losing sequences are endless.  Winning sequences include:


Theoretically, you  never  have to concede defeat:  Even if you have already flipped 1000 tails, there's a very slim chance that you will flip 1001 heads in a row, starting with the next throw.  In practice, it would be silly not to give up after accumulating  30  tails or so...  As such an artificial bound tends to infinity, the probability of winning tends rapidly to a value of  0.7112119...

Let's analyze the situation without assuming that the coin is a fair one:  p  is the probability of heads and  q = 1-p  is the probability of tails.

We may bypass the elementary considerations of the above introduction by dealing directly with the uncountable set of infinite sequences of flips.  The subset of all such sequences sharing a prescribed prefix consisting of  n  heads and  m  tails has a  probability measure  equal to  pqm.

Let's call  P(n)  the probability of the set of all strings that start with a winning string with  n  tails stripped from its  n+1  trailing heads.  Such a string with  n+1  tails is uniquely obtained by appending to a string of the same type with  n  tails a string consisting of at most n consecutive heads, followed by one final tails.  Therefore:

P(n+1)   =   P(n) q   å   pi     =   P(n) ( 1-pn+1 )

So,  P(0) = 1,   P(1)  =  (1-p),   P(2)  =  (1-p)(1-p2 ),   etc.

The probability of the set of winning strings with  n  tails being   pn+1 P(n)   we obtain the following expression for the overall probability of winning:

¥ n
f (p)   =   p  +    å   pn+1  Õ   ( 1-pk )
n=1 k=1

The above probability of winning may also be expressed as:

f (p)   =   p [ 1 + p (1-p) [ 1 + p (1-p2 ) [ 1 + p (1-p3 ) [ 1 + p (1-p4 ) [ ... ] ]]]]

As advertised, for a fair coin  (p = ½)  that expression is equal to:

f (½)   =   0.711211904913397578721100278070769219911088...

On 2010-09-25, Michael Wales wrote:   [ edited summary ]
This is the greatest site ever.   I've forwarded your link to others to whom I've asked the same question.

Thanks for the kind words, Michael.

The above function  f  has an interesting expansion as a  power series :

f (x)   =   x + x2 - x5 - x7 + x12 + x15 - x22 - x26 + x35 + x40 - x51 - x57 + x70 + x77 - x92 - x100 + x117 + x126 - x145 - x155 + x176 + x187 - x210 - x222 + x247 + x260 - ...

The exponents of  x  are the  generalized pentagonal numbers  (A001318).  The terms are grouped by pairs, with alternating signs.  More precisely:

f (x)   =     å   x (2k+1)(3k+1)  +  x (2k+1)(3k+2)  -  x (k+1)(6k+5)  -  x (k+1)(6k+7)

That formula yields a fast way to compute  f (x)  and a peculiar decimal expansion for  f (0.1)  that features only 4 digits  (0, 1, 8 and 9)  namely:


Incidentally:   ½  =  f ( 0.3705997158021455417170957127716712...)

( Brad of Schaumburg, IL. 2000-11-03)
How many ways can you make change for a dollar?

If you only allow quarters, dimes, nickels and pennies, the answer is 242. If you also allow dollar coins and half-dollars, the answer is 293.

Note that the number of ways you can obtain a sum of 100 by adding 25,10,5, and 1 is the coefficient of x100 in the expression:

(1 + x25 + x50 +...) (1 + x10 + x20 + x30 +...) (1 + x5 + x10 + x15 +...) (1 + x + x2 + x3 + x4 +...)

also equal to: 1 / [ (1-x25)(1-x10)(1-x5)(1-x) ]

The practical way to do it is via a small computer program. In BASIC, this may be something as simple as the following (which prints "242"):

s = 0
FOR q = 100 TO 0 STEP -25
  FOR d = q TO 0 STEP -10
    FOR n = d TO 0 STEP -5
      s = s + 1

It's easy to change the above to account for half-dollars (the result is 292).  Add the unique way corresponding to a single dollar coin for a total of 293.

The above simple-minded program can be made far more efficient if we notice that there are  1+floor(p/5)  ways to make change for an amount of  p  cents with nickel and pennies only  (that's what the inner loop computes).  The same idea is easy to extend to get rid of the next innermost loop by counting the ways to make change with pennies, nickels and dimes; there are  (1+d)(1-d+n)  such ways, where n=floor(p/5) and d=floor(p/10).  The following program is indeed much faster for large values of p (well beyond p=100):

INPUT "Amount in cents p=";p
FOR j = p TO 0 STEP -25
  s = s + (1 + INT(j/5))*(1 - INT(j/5) + INT(j/10))
PRINT s;"ways; with pennies, nickels, dimes & quarters"

It's more difficult to get rid of this "quarter counting" loop with a closed formula, but it can be done (see below).  When we are faced with a large number of weird denominations it may not be practical to derive such closed form formulas.  On the other hand, programs like the above tend to be too slow  [their running time is roughly proportional to the result, which grows exponentially with the number of denominations].  A more efficient approach is based on the actual computation of the product of the polynomials described above.  This may be done with a number of additions which is roughly proportional to the number of denominations, and also to the square of the amount p...

Given below is a BASIC implementation of that approach.  As is, the program is written for pennies, nickels, dimes and quarters, but you may have different and/or additional denominations by changing only the first line of the program to define other coins (in terms of the number of pennies they stand for).  Two polynomials are used in the form of arrays (A and B).  For each new coin denomination, Polynomial B is updated to reflect the number of ways to make change with such coins and/or the previously considered ones, knowing that the (previously computed) Polynomial A tells how many ways change could be made without the new denomination...  At the end, Array B contains all the answers for any amount p, up to the built-in limit M defined at the beginning of the program (M=200 is just an example here). It's then a simple matter to display interactively the answer B(p) for any desired amount p:

N=3: DIM c(N): c(1) = 5: c(2) = 10: c(3) = 25
M=200: DIM A(M),B(M)
FOR i = 0 TO M: B(i) = 1: NEXT i

FOR d = 1 TO N
  FOR i = 0 TO M: A(i) = B(i): B(i) = 0: NEXT i
  FOR j = 0 TO M STEP c(d)
    FOR k = 0 TO M-j
      B(j+k) = B(j+k) + A(k)
    NEXT k
  NEXT j

  INPUT "Amount in cents p=";p
  PRINT B(p); "ways!"
Number of ways to make change with given coin denominations.
Amount1,51,5,101,5,10,25 1,5,10,25,501,5,10,25,50,100
100 21121242292293
200 41441146324352728
500 1012601190065957698411
1000 201102011425118014512103596
2000 4014040111030211171210153995291
5000 100125100116892551 4326992514587118976
10000 20011002001134235101 6794128501139946140451
(n+1)(n-d+1)(d+1)See formulas below...

Making change with US coins:  Explicit formulas.

Let p be the amount in cents (¢) for which we have to make change.  Let n be floor(p/5) which is the number of whole nickels (5¢) contained in that.  Similarly, we'll call d,q,h and w the numbers of dimes (10¢) quarters (25¢) half-dollars (50¢) and whole dollars (100¢) in the amount p.

The number of ways to make change using only pennies and nickels is simply n+1, because all you have to do is decide on a number of nickels from 0 to n.

If you're allowed dimes as well, you first choose a number i of dimes from 0 to d, and are left with (p-10i) cents for which you must give change using only pennies and nickels.  Since floor((p-10i)/5) is n-2i, you have n-2i+1 ways to do that.  The total number of ways to make change is thus the sum of (n-2i+1) when i goes from 1 to d, which works out to be:

(n-d+1)(d+1)   ways to give change for p cents
using pennies, nickels & dimes,
where   n = floor(p/5)   and   d = floor(p/10)

The situation is more complicated when quarters are allowed as well:

If we use an even number 2j of quarters, we have to make change for p-50j cents without quarters and have (n-d-5j+1)(d-5j+1) ways to do so...

With an odd number 2j+1 of quarters, we're left with (p-50j-25) pennies.  Since floor((p-50j-25)/10) can be shown to be equal to (n-d-5j-3), we have (d-5j-1)(n-d-5j-2)  ways to complete the transaction.

When q = 2h, we sum up the first expression for j going from 0 to h and the second one for j between 0 and h-1.  When q = 2h+1, on the other hand, we sum the second one also up to j = h, which gives formally an additional term equal to (d-5h-1)(n-d-5h-2) . Thus, we obtain an expression valid when q is either 2h or 2h+1 by adding (q-2h)(d-5h-1)(n-d-5h-2) to the expression obtained for q = 2h.  The formidable result shows that there are  133333423333351000001 ways to make change for a  million dollars  in pennies, nickels, dimes and quarters:

(h / 6)  [ 100 h + 6 d (2n-2d-1) - 15 h (2n-1) - 7 ]
+   (n-d+1) (d+1)   +   (q-2h) (d-5h-1) (n-d-5h-2)

Eliminating h  [via  q = 2h + (q mod 2) ]  we obtain the following expression.  (The last term is  0  or  ±1/8  and may thus be discarded in a rounded result.)

(q / 24)  [ 50 q + 12 d (2n-2d-1) - 15 q (2n-1) - 14 ]
+   (n-d+1) (d+1)   +   (q mod 2) [ 2(n-2d) - 1 ]  / 8

NOTE:   This formula was worked out and posted here on February 8, 2004.  If you're aware of an earlier source, please tell us, and we'll give proper credit... Otherwise, please give us credit:  The permanent link to (the latter part of) this article is http://www.numericana.com/answer/counting.htm#uschange.  Thanks.

(2009-04-22)   British Currency   (Project Euler, Problem 31)
Using 8 denominations of  1, 2, 5, 10, 20, 50, 100 and 200 pennies, how many ways are there to make change for  N  pennies?

A simple efficient program can be quickly designed as above which gives foolproof answers when  N  is less that the size  M  of a reasonable array.

Mint=7: DIM c(Mint)
c(1)=2: c(2)=5: c(3)=10: c(4)=20: c(5)=50
c(6)=100: c(7)=200
M=200: DIM A(M),B(M)
FOR i = 0 TO M: B(i) = 1: NEXT i

FOR d = 1 TO Mint
  FOR i = 0 TO M: A(i) = B(i): B(i) = 0: NEXT i
  FOR j = 0 TO M STEP c(d)
    FOR k = 0 TO M-j
      B(j+k) = B(j+k) + A(k)
    NEXT k
  NEXT j

  INPUT "Amount in pence N=";N
  PRINT B(N); "ways!"

However, mathematics is the art of patterns and we may want to go after the same type of closed formulas which were derived for American money at the end of the previous section.  That would allow values of  N  in the zillions.

To enforce a better computational discipline, we shall consider that the supply of various coins is  limited.  Let's introduce the notation:

Ni   =   min { floor ( N / i ) ,  number of available i-coins }

Assuming N1 = N, there's just  one  way to make change for N with pennies only, isn't there?  (With N1<N, there would be no way to make change.)

Using 1 and 2 denominations  (pennies and tuppences)  you may choose to use from 0 to N2 tuppences and the balance in pennies only.  There are  1+N2 ways to do so  (still assuming N1=N).

If you also have shillings  (5p, five pence coins)  you may use any number  s  of them, between 0 and N5.  You are then left with an amount  N-s  for which the previous formula applies.  So, the total number of possibilities is

N5     1 +  min ( N2 , floor((N-s)/2) )
s = 0

 Come back later, we're
 still working on this one...

(L. S. of Canada. 2000-11-20)
How many 5-digit numbers have their digits in a decreasing sequence?

There are 252 decreasing sequences of 5 digits (and 2002 nonincreasing ones).

To build such a sequence, you must choose 5 digits among the 10 possible ones.  Once you've done this, you've got no choice but to lay them out in decreasing order.  Therefore, this can be done in C(10,5)=252 ways.

If you were interested in a nonincreasing sequence instead, you would allow several adjacent copies of the same digit.  This is only slightly more difficult to count; there are C(14,5)=2002 possibilities.

Why is that?  Well, consider a row of 9+5=14 squares of which 5 are "occupied" and 9 are "empty".  In each occupied square, just write the total number of empty squares to its right.  The 5 numbers you wrote are between 0 and 9 and their sequence is nonincreasing.  Conversely, each nonincreasing sequence of 5 digits may occupy (under the above rules) 5 of 14 such squares in a unique way; place them right to left to see that you have no choice in the matter.  Therefore, there are just as many nonincreasing sequences of 5 digits as there are ways to pick 5 squares among 5+9.  That's C(14,5)=2002.

(C.C. of Baton Rouge, LA. 2000-10-08)
1. A deck of 52 playing cards are divided into 4 piles of 13 cards each. Find the probability that each pile has exactly one ace.
2. If a box contains 75 good IC (integrated circuit) chips and 25 bad chips, and 10 are selected at random, find the probability that there are exactly 2 defective chips in the 10 selected.
3. An airline overbooks its flight with a capacity of 350 by 20 (i.e. it makes 370 reservations). If the probability of a passenger not showing up is 5 percent, what are the chances that not every passenger that shows up will get a seat.

1. There is a total of C(52,4) equiprobable ways of choosing 4 positions for the aces in the (ordered) piles and 134 ways of choosing one position in each pile. The probability that each pile has one ace is thus 134/C(52,4)=2197/20825 or about 0.105498199... just under 10.55%.

2. C(100,10) ways of picking 10 ICs and C(25,2)C(75,8) ways of picking 2 bad ones and 8 good ones, for a probability of C(25,2)C(75,8)/C(100,10)=11002861125/37631107514 or about 29.24%.

3. The probability of exactly k passengers showing up is C(370,k)(0.95)k(0.05)370-k. Add these up from k=351 to k=370 and you'll get the probability that more than 350 passengers will show up, which is just under 60.7264% (0.6072639447).

( Erin of Edinboro, PA. 2000-10-24)
Find the probability of getting two face cards (jack, queen, or king) when two cards are drawn from the deck without replacement.

11 / 221   is your answer:

In a 52 card deck, there are 12 face cards. There are C(52,2) equiprobable combinations of 2 cards among which only C(12,2) are pairs of face cards. The probability of getting a pair of face cards is thus C(12,2)/C(52,2) which is 12´11/(52´51) = 11/221, or about 0.04977...  In other words, there are "11 chances in 221" (exact result); a probability of about 4.977%.

If you were to put the first card back in the deck before the second draw, the probability of getting a second face card would be higher, since you'd improve the chances of getting a "good" card on the second draw whenever you were successful the first time around.  (The result would be the probability of getting a face card twice in two independent events, namely (3/13)2, or about 5.325%.)

(J.O. of United Kingdom. 2000-29-09)
How many rectangles can be found in a square grid of 14 x 14 squares?

11025 for a square grid of side N=14. This has been asked/answered many times for other values of N (N=8 for a chessboard).

My solution is to count the number of possible diagonals of such rectangles: Each diagonal is uniquely determined by a pair of extremities. The number of such pairs is the product of (N+1)2 (choices for the first point) and N2 (choices for the second point not on the row or the column of the first) divided by 2 (because there are two ordered pairs for every unordered pair). As there are 2 diagonals per rectangle, the above product is simply twice the number of rectangles. Thus, the number of rectangles is N2(N+1)2/4, which is 11025 for N=14 (and 1296 for N=8, by the way).

(Katie of Morton, PA. 2000-10-15)   A000330
What is the total numbers of squares that are in a checkerboard?
I'm not looking for the answer 64, but more than that.
 16 ways to choose a 5 by 5 
 square on a chessboard

The answer is 204 for a regular 8 by 8 checkerboard.  It's N(N+1)(2N+1)/6 for an N by N checkerboard.  Here's one way to prove this:

A square of size P is uniquely determined by its lower left corner, which may be picked anywhere on the (N-P+1) by (N-P+1) grid at the lower left of the chessboard. The total number of squares is therefore N2+(N-1)2+(N-2)2+...+22+12, which is equal to N(N+1)(2N+1)/6. Namely:

  • 204 for a regular chessboard (N=8),
  • 285 for a shogi board (N=9),
  • 385 for the board in international checkers (N=10),
  • 1240 for a Scrabble(tm) board (N=15),
  • 2109 for a go board (19 lines in either direction; N=18).

Philip Brocoum  (2007-03-10; e-mail)   Screaming Circles
10 people stand in a circle and look at the floor.  On the count of three, everyone looks up to somebody else.  Each person must choose to look at one of 3 people: The person to the right, the person to the left, the person directly across the circle.  When two people make eye contact, both scream.  What's the probability that no one will scream?

Philip Brocoum presents this as a warm-up exercise which was actually practised by his former improv troupe at MIT, until nobody screamed.  (2006-01-26)

As each of the 10 people has 3 choices, there are only  310  =  59049  equiprobable possibilities.  By direct counting, only 4406 of these involve no direct "eye contact".  So, the probability (p) that no one will scream is about  7.46 %.  The  expected  number (1/p) of times it would take to repeat the experience until nobody screams is thus slightly more than 13.4.

For an even number (n) of people, there are  3n  scenarios.  The number of those for which nobody screams is given by the second row of the following table, whose third row gives the expected number of tries needed to obtain a silent one.

n 4681012141618
Silent Cases 30156826440623562 1261046750743614142
1 / p 37.963.8107.2

Brocoum's problem can be usefully generalized, to  Hamiltonian cubic graphs.  For Brocoum's original poser,  Max Alekseyev  found (on 2008-06-13)  that the number of silent configurations for a circle of  2n  people (A141221)  is given by the following recurrence relation, for  n>1.  (see full proof elsewhere).

an+4   =   8 an+3  -  16 an+2  +  10 an+1  -  an

It's  also  fully defined by a  third order  recurrence  (cf. A141385) :

an+3   =   7 an+2  -  9 an+1  +  an  -  2
starting with:     a2 = 30   ;     a3 = 156   ;     a4 = 826

The sequence is best started at  n=2  (a circle of 4 people)  as the rules of the game do not really apply to a "circle" of only two people.  (Alekseyev's analysis does not apply to a circle of less than 4 people either.)

What's the average distance between two random points on a segment?

In all questions that involve random variables which may assume a continuous range of values, we must carefully specify the probability distribution involved.  This is often more delicate than in the discrete case. Here, it suffices to say that if 0<a<b<1, a "random number" between 0 and 1 falls between a and b with probability b-a.  This is not a statement to be proven, it's a description of the Lebesgue probability measure on the unit segment (the so-called uniform probability distribution), which we choose to assume.  With another choice of distribution, we would obtain other results.  That's all.

Well, with this distribution, the average distance from a random point in a segment to either of its extremities is half the length of the segment.  [The reader is encouraged to prove this.]  If we consider a fixed point x on the unit segment [0,1], a random point y will be less than x with probability x, and it will be more than x with probability (1-x).  The distance between x and y averages x/2 in the former case, and (1-x)/2 in the latter.  The overall average distance to point x is obtained as the weighted average of the two cases (each case is weighted according to its own probability).  It is thus equal to:

x2/2 + (1-x)2/2   =   x2 - x + 1/2   =   d/dx [ x3/3 - x2/2 + x/2 ]

If x is a random number uniformly distributed between 0 and 1, the average of the above quantity is equal to its integral from 0 to 1, namely 1/3.  Therefore:

The average distance between two points [uniformly] picked at
random on a segment is one third the length of the segment.

A nice introduction to continuous random variables and/or elementary calculus...

It's much more difficult to work out the average distance between two random points on a unit squaregiven by Mark R. Diamond (2001-12-11):

[2 + Ö2 + 5 ln(1+Ö2) ]  / 15   =   0.52140543316472067833...

Video :   Distance between Two Random Points in a Square   by  Presh Talwalkar  (2016-07-03).

Anonymous Google query  (2006-10-08)
Average distance between two points on the surface of a sphere.

Since all points on the surface of a sphere of radius  R  are equivalent, the average distance between two random points is the same as the average distance from a fixed point to a random one  (assuming, of course, that the probability of a set is proportional to its spherical area).

Average Distance Along a Great Circle :

p R / 2

HINT :  That result is obtained immediately by noticing a symmetry about the equator  (at a distance pR/2 from either pole)  concerning distances to the pole.

Average Euclidean Distance :

4 R / 3

Distances in Bounded Regions

Steve Blue442  (2011-07-26, e-mail)   Buffon's Needle  redux
Jumping a distance  D < Ö2  from a random point inside a unit square, what's the probability of remaining inside it?

Let's assume that the random starting position is uniformly distributed within the unit square.  This is to say that the probability that the random point is within some patch is proportional to the  area  of that patch  (actually, the probability is equal to the area since the total area of the square is unity).

Let's first assume that we are only allowed to jump along a direction at an angle  q  from a diagonal  (WLG, q is between  0  and  45°).  A jump of length  D  stays within the unit square when the starting point is in a rectangle whose area is:

( 1 - | D cos q | ) ( 1 - | D sin q | )

That quantity is thus the probability that we stay within the square when the direction of the jump is prescribed.  If the direction is random  (isotropically distributed)  the desired probability is the average of the above, namely:

2 ù  p/2
p 0

Buffon's Needle :

 Come back later, we're
 still working on this one...

Distances in Bounded Regions

visits since Dec. 6, 2000
 (c) Copyright 2000-2019, Gerard P. Michon, Ph.D.