Probability

Axioms of Probability

Wang Fei

matwf@[Link]

Department of Mathematics
Office: S17-06-16
Tel: 6516-2937

Sample Space and Events 2

Sample Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Operations on Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Axioms of Probability 11
Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Axioms of Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Some Simple Propositions 16

Properties of Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Inclusion-Exclusion Identity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Sample Spaces Having Equally Likely Outcomes 25

Equally Likely Outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Probability as a Continuous Set Function 37

Monotone Sequence of Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Limit of Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Probability as a Measure of Belief 41

Measure of Belief. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

1
Sample Space and Events

Sample Space
✔ Definition. Consider an experiment whose outcome is not predictable.
✘ The set of all possible outcomes is called the sample space.
✓ It is usually denoted by S, unless stated otherwise.
✔ Examples.
✘ Determine the sex of a newborn child.
✓ S = {boy, girl}.
✘ Flip two coins. Use H for head and T for tail.
✓ S = {(H, H), (H, T ), (T, H), (H, H)}.
✘ The lifetime of a transistor (measured in hours).
✓ S = {t | 0 ≤ t < ∞} = [0, ∞).

Events
✔ Definition. Any E subset of the sample space S is an event.
✘ If the outcome is contained in E, then we say E occurs.
✔ Examples.
✘ S is itself an event, the sure event; ∅ is the null event.
✘ Determine the sex of a new born child.
✓ E = {girl} is the event that the child is a girl.
✘ Flip two coins.
✓ E = {(H, H), (H, T )} is that a head appears in the 1st coin.
✘ The lifetime of a transistor (in hours).
✓ E = {t | 0 ≤ t ≤ 5} = [0, 5] is the event that the transistor does not last longer
than 5 hours.

2
Operations on Events
✔ Operations on Events are precisely operations on sets.
✘ Let E and F be two events of a sample space S.
✓ The union E ∪ F consists all outcomes in E or in F or in both.
S
E F

✓ It is clear that E ∪ F = F ∪ E.

Operations on Events
✔ Operations on Events are precisely operations on sets.
✘ Let E and F be two events of a sample space S.
✓ The intersection E ∩ F (EF ) consists all outcomes in both E, F .

S
E F

✓ It is clear that EF = F E.
✓ If EF = ∅, then E and F are called mutually exclusive.

3
Operations on Events
✔ Operations on Events are precisely operations on sets.
✘ Let E be an event of the sample space S.
✓ The complement E c consists all outcomes (in S) not in E.

S
E Ec

✓ It is clear that S c = ∅, ∅c = S, (E c )c = E.

Operations on Events
✔ Operations on Events are precisely operations on sets.
✘ Let E1 , E2 , . . . be a sequence of events in the sample space S.
S∞
✓ Their union, denoted by n=1 En ,
✗consists of all outcomes which are in at least one of En .
T∞
✓ Their intersection, denoted by n=1 En ,
✗ consists of all outcomes which are in every En .
✔ Examples.
Let En = {1, 2, 3, . . . , n}. Then ∞
S
✘ n=1 En = N.

✘ Let Fn = x | − n1 < x < n1 = − n1 , n1 . Then n=1 Fn = {0}.


T∞

4
Properties
✔ Theorem. Let E, F, G be events in a sample space S.
✘ Commutative Laws:
✓ E ∪ F = F ∪ E, EF = F E.
✘ Associative Laws:
✓ (E ∪ F ) ∪ G = E ∪ (F ∪ G), (EF )G = E(F G).
✘ Distributive Laws:
✓ (E ∪ F )G = EG ∪ F G, (EF ) ∪ G = (EG) ∪ (F G).
✘ De Morgan’s Laws:
✓ (E ∪ F )c = E c F c , (EF )c = E c ∪ F c .

Properties
✔ Theorem. Let E1 , E2 , . . . and F be events in a sample space S.
✘ Distributive Laws:
S∞ S∞ T∞ T∞
✓ ( n=1 En ) ∩ F = n=1 (En ∩ F ); ( n=1 En ) ∪ F = n=1 (En ∪ F ).
✘ De Morgan’s Laws:
✓ ( n=1 En )c = n=1 Enc , ( n=1 En )c = n=1 Enc .
S∞ T∞ T∞ S∞

Proof of ( ∞ c
T S∞ c
n=1 En ) = n=1 En . (Others are left as exercises.)

✘ Let x be an outcome in ( n=1 En )c . Then x is not in n=1 En .

T∞ T∞

✓ x is not in some En , i.e., x is in some Enc ; so x ∈ n=1 Enc .

S∞
S∞
✘ Let x be an outcome in n=1 Enc . Then x is in some Enc .
T∞ S∞
✓ x is not in some En , i.e., x is not in n=1 En ; so x is in ( n=1 En )c .

5
Axioms of Probability

Probability
✔ Flip a fair coin n times. Then S = {H, T }.
✘ Let E = {H} be the event of having a head.
Flip the coin n times, and let n(E) be the number of times that E occurs.
✘ If n is large, it is expected that n(E) is approximately 12 n.
n(E) 1
✓ In other words, lim = .
n→∞ n 2
This limit is called the probability of E, denoted by P (E).
✔ In general, let E be any event of an experiment.
✘ Let n(E) be the number of times that E occurs in the first n repetitions of the
experiment.
n(E)
✓ Theprobability of E is P (E) = lim , if this limit exists.
n→∞ n

Probability
✔ Let S be the sample space of an experiment. Assume that for any event E,
n(E)
✘ P (E) = lim is well-defined.
n→∞ n
Then E must satisfy some properties.
1. Note that 0 ≤ n(E) ≤ n. So 0 ≤ n(E)/n ≤ 1.
✓ If P (E) = lim n(E)/n exists, we must have 0 ≤ P (E) ≤ 1.
n→∞

2. The sample space S consists of all outcomes. So n(S) = n.

✓ It follows that n(S)/n = 1 and thus P (S) = 1.
3. Let E and F be mutually exclusive events, i.e., EF = ∅.
✓ Then n(E ∪ F ) = n(E) + n(F ).
✓ It follows that P (E ∪ F ) = P (E) + P (F ).

6
Probability
✔ Let S be the sample space of an experiment. Assume that for any event E,
n(E)
✘ P (E) = lim is well-defined.
n→∞ n
Then E must satisfy some properties.
1. Note that 0 ≤ n(E) ≤ n. So 0 ≤ n(E)/n ≤ 1.
✓ If P (E) = lim n(E)/n exists, we must have 0 ≤ P (E) ≤ 1.
n→∞

2. The sample space S consists of all outcomes. So n(S) = n.

✓ It follows that n(S)/n = 1 and thus P (S) = 1.
3. Let E1 , E2 , . . . be mutually exclusive events, i.e., Ei Ej = ∅ for i 6= j.
S∞ P∞
✓ Then n( n=1 En ) = n=1 n(E).
S∞ P∞
✓ It follows that P ( n=1 En ) = n=1 P (En ).

Axioms of Probability
✔ Axioms of Probability. Let S be the sample space of an experiment.
✘ Suppose that a number P (E) is defined for every event E of S such that
1. 0 ≤ P (E) ≤ 1.
2. P (S) = 1.
3. For any sequence of mutually exclusive events E1 , E2 , . . . ,
S∞ P∞
✗ P ( i=1 Ei ) = i=1 P (Ei ).
Then P (E) is called the probability of the event E.
Remarks. It is natural to see the rationality of finite additivity:
Sn Pn
✘ P ( i=1 Ei ) = i=1 P (Ei ) for mutually exclusive events E1 , . . . , En .
However, countable additivity (axiom 3) is necessary to deal with sample space with infinite
outcomes.

7
Some Simple Propositions

Properties of Probability
✔ Let ∅ = E1 = E2 = E3 = · · · . Then E1 , E2 , . . . are mutually exclusive.
S∞ P∞ P∞
✘ P ( i=1 Ei ) = i=1 P (Ei ) reduces to P (∅) = i=1 P (∅).
✓ It follows that P (∅) = 0.
✔ Let E1 , . . . , En be mutually exclusive and ∅ = En+1 = En+2 = · · · .
S∞ Sn
✘ P ( i=1 Ei ) = P ( i=1 Ei ).
P∞ Pn P∞ Pn
✘ i=1 P (Ei ) = i=1 P (Ei ) + i=n+1 P (∅) = i=1 P (Ei ).
Sn Pn
✓ It follows that P ( i=1 Ei ) = i=1 P (Ei ).
✔ Example. Flip a coin so that a head is twice as likely to appear as a tail.
✘ P ({H}) = 2P ({T }) and P ({H}) + P ({T }) = P ({H, T }) = 1.
2
✓ Then P ({H}) = 3
and P ({T }) = 13 .

Properties of Probability
✔ Note that E and E c are mutually exclusive events.
✘ P (E) + P (E c ) = P (E ∪ E c ) = P (S) = 1.
✓ It follows that P (E c ) = 1 − P (E).
✔ Suppose that E ⊆ F .
✘ The event that F occurs but E does not occur is F \ E = F E c .
✓ E ∩ F E c = ∅ and E ∪ F E c = F .
✘ P (E) + P (F E c ) = P (F ).
✓ It follows that P (E) ≤ P (F ).

8
Properties of Probability
✔ Let E and F be any two events.

E F
c c
EF EF FE

✘ Then EF c , EF and F E c are mutually exclusive with union E ∪ F .

P (E ∪ F ) = P (EF c ) + P (EF ) + P (F E c )
= [P (EF c ) + P (EF )] + [P (EF ) + P (F E c)] − P (EF )
= P (E) + P (F ) − P (EF ).

Properties of Probability
✔ Propositions. Let E1 , . . . , En and E, F be any events in a sample space.
✘ P (∅) = 1.
✘ If E1 , . . . , En are mutually exclusive,
Sn Pn
✓ P ( i=1 Ei ) = i=1 P (Ei ).
✘ P (E c ) = 1 − P (E).
✘ If E ⊆ F , then P (E) ≤ P (F ).
✘ Inclusion-Exclusion Identity:
✓ P (E ∪ F ) = P (E) + P (F ) − P (EF ).

9
Examples
✔ J likes 1st book with probability 0.5, the 2nd book with probability 0.4, and both books
with probability 0.3.
✘ Find the probability that J likes neither book?
Solution. Let Bi be the probability that J likes the ith book.
✘ It is given that P (B1) = 0.5, P (B2 ) = 0.4, P (B1 B2 ) = 0.3.
✓ P (B1 ∪ B2 ) = P (B1 ) + P (B2 ) − P (B1 B2 ) = 0.6.
✘ It is required to find the probability of B1c B2c .
✓ By De Morgan’s law, B1c B2c = (B1 ∪ B2 )c .
✓ P (B1c B2c ) = P ((B1 ∪ B2 )c ) = 1 − P (B1 ∪ B2 ) = 0.4.

Inclusion-Exclusion Identity
✔ Let E, F and G be any three events.

G 3

E 5 6 F
7

1 4 2

P (E ∪ F ∪ G) = {1, 4, 5, 7} + {2, 4, 6, 7} + {3, 5, 6, 7}

− {4, 7} − {5, 7} − {6, 7} + {7}
= P (E) + P (F ) + P (G)
−P (EF ) − P (EG) − P (F G) + P (EF G).

10
Inclusion-Exclusion Identity
✔ Inclusion-Exclusion Identity.
✘ P (E ∪ F ) = P (E) + P (F ) − P (EF ).
✘ P (E ∪ F ∪ G) = P (E) + P (F ) + P (F )
−P (EF ) − P (EG) − P (F G) + P (EF G).
✘ P (E1 ∪ E2 ∪ E3 ∪ E4 ) = P (E1 ) + P (E2 ) + P (E3 ) + P (E4 )
−P (E1 E2 ) − P (E1 E3 ) − P (E1 E4 ) − P (E2 E3 ) − P (E2 E4 ) − P (E3 E4 )
+P (E1 E2 E3 ) + P (E1 E2 E4 ) + P (E1 E3 E4 ) + P (E2 E3 E4 )
−P (E1 E2 E3 E4 ).
Sn n
(−1)r+1
P P
✔ More generally, P ( i=1 Ei ) = P (Ei1 · · · Eir ).
r=1 i1 <···<ir
n
P P P
P (E1 ∪ · · · ∪ En ) = P (Ei ) − P (Ei1 Ei2 ) + P (Ei1 Ei2 Ei3 )
i=1 i1 <i2 i1 <i2 <i3
n+1
− · · · + (−1) P (E1 E2 · · · En ).

Inclusion-Exclusion Identity
✔ Let E1 , E2 , . . . , En be events.
✘ Let Fi consists of all outcomes in Ei but not in E1 , . . . , Ei−1 .
Sn Sn
✓ F1 , . . . , Fn are mutually exclusive and i=1 Fi = i=1 Ei .
Then P ( ni=1 Ei ) = P ( ni=1 Fi ) = ni=1 P (Fi ) ≤ ni=1 P (Ei ).
S S P P

Boole’s Inequality: P ( ni=1 Ei ) ≤ ni=1 P (Ei ).

S P

✘ The equality holds if E1 , . . . , En are mutually exclusive.

More generally, we have Inclusion-Exclusion Inequalities:
Sn Pn
✘ P ( i=1 Ei ) ≤ i=1 P (Ei ).
Sn Pn P
✘ P ( i=1 Ei ) ≥ i=1 P (Ei ) − i<j P (Ei Ej ).
Sn Pn P P
✘ P ( i=1 Ei ) ≤ i=1 P (Ei ) − i<j P (Ei Ej ) + i<j<k P (Ei Ej Ek ).

11
Sample Spaces Having Equally Likely Outcomes

Equally Likely Outcomes

✔ Flip a fair coin.
✘ Assume that P ({H}) = P ({T }).
✔ Toss a fair die.
✘ Assume that P ({1}) = P ({2}) = · · · = P ({5}) = P ({6}).
✔ Draw a card from a deck of 52 cards.
✘ Assume that P ({Ace of Heats}) = · · · = P ({King of Spades}).
✔ Suppose the sample space S of an experiment is finite: S = {x1 , . . . , xN }.
✘ Assume all outcomes in the sample space are equally likely to occur,
✓ i.e., P ({x1 }) = P ({x2 }) = · · · = P ({xN }).
1
Then P ({xi }) = for each i = 1, . . . , N.
N

Equally Likely Outcomes

✔ Suppose the sample space S of an experiment is finite such that
✘ all outcomes in S are equally likely to occur.
|E|
For any event E, P (E) = , where |E| is the number of outcomes in E.
|S|
✔ Example. Toss two fair dice.
✘ Find the probability that the sum is 7.
Solution. The sample space is S = {(i, j) | i, j = 1, . . . , 6}.
✘ Let E be the event that the sum is 7.
✓ E = {(1, 6), (2, 5), (3, 4), (4, 3), (5, 2), (6, 1)}.
|E| 6 1
Then P (E) = = = .
|S| 36 6

12
Examples
✔ 3 balls are randomly drawn from a bowl with 6 white and 5 black balls.
✘ Find the probability of drawing 1 white and 2 black balls.
Solution. Suppose that the order is relevant.
✘ There are 11 × 10 × 9 = 990 ways of drawing 3 from 11 balls.
To select 1 white and 2 black balls,
(i) (W, B, B): there are 6 × 5 × 4 = 120 ways.
(ii) (B, W, B): there are 5 × 6 × 4 = 120 ways.
(iii) (B, B, W ): there are 5 × 4 × 6 = 120 ways.
✓ In total: 120 + 120 + 120 = 360 ways.
360 4
Therefore, the desired probability is = .
990 11

Examples
✔ 3 balls are randomly drawn from a bowl with 6 white and 5 black balls.
✘ Find the probability of drawing 1 white and 2 black balls.
Solution. Suppose that the order is irrelevant.
✘ There are 11


3
= 165 ways of drawing 3 balls from 11 balls.
To select 1 from 6 white balls and 2 from 5 black balls,
✘ there are 61 × 52 = 60 ways.

60 4
Therefore, the desired probability is = .
165 11

13
Examples
✔ A committee of 5 is randomly selected from 6 men and 9 women.
✘ Find the probability of having 3 men and 2 women in the committee.
Solution. There are 15


5
= 3003 ways to form the committee.
✘ To select 3 from 6 men, and 2 from 9 women,
✓ there are 63 × 92 = 720 ways.

720 240
The desired probability is 3003
= 1001
.
✔ m men and n women are arranged so that any ordering is equally likely.
✘ If only genders are recorded, consider any of the (m + n)! orderings.
✓Any permutation of men among themselves and of women among themselves
doesn’t change the order of genders — m!n! in all.
m!n!
= 1/ m+n


✘ The probability for each ordering of genders is (m+n)! m
.

Examples
✔ An urn contains n balls, one of which is special.
✘ Balls are withdrawn one at a time such that each selection is equally likely to be any of
the balls that remain.
✓ Find the probability that the ith ball is special.
Solution. There are |S| = n! equally likely outcomes.
✘ Let Ai be the event that the ith ball is special.
✓ |Ai | = (n − 1) · · · [n − (i − 1)] × 1 × (n − i)[n − (i + 1)] · · · 2 · 1.
|Ai | (n − 1)! 1
✓ P (Ai ) = = = .
|S| n! n
In other words, the special ball is equally likely to appear in any position of the selection.

14
Examples
✔ A deck of 52 cards is shuffled.
✘ Cards are turned up one at a time until the first ace appears.
(a) Find the probability that the next card is the ace of spades.
(b) Find the probability that the next card is the two of clubs.
Solution. Note that any of 52! ordering is equally likely.
✘ Arrange other 51 cards and then insert the ace of spades.
✓ There is one position for the ace of spades after the first ace.
So there are 51! orderings so that ace of spades is after the first ace.
51! 1
✓ The desired probability is = .
52! 52
For any particular card (including the ace of spades),
✘ the probability that it is next to the first ace is 1/52.

Examples
✔ Matching Problem. Each of N men throws his hat.
✘ The hats are mixed and then each man randomly select one hat.
Find the probability that none of the men selects his own hat.
✔ Solution. Let N = 3.
✘ Let Ei denote the event that the ith man selects his own hat.
✓ |Ei | = 2!; |Ei Ej | = 1!; |E1 E2 E3 | = 0!.

P (E1 ∪ E2 ∪ E3 ) = P (E1 ) + P (E2 ) + P (E3 )

− P (E1 E2 ) − P (E1 E3 ) − P (E2 E3 ) + P (E1 E2 E3 )
2! 1! 0!
= 3· 3!
−3· 3!
+ 3!
= 23 .
The desired probability is P (E1c E2c E3c ) = 1 − P (E1 ∪ E2 ∪ E3 ) = 13 .

15
Examples
✔ Matching Problem. Each of N men throws his hat.
✘ The hats are mixed and then each man randomly select one hat.
Find the probability that none of the men selects his own hat.
✔ Solution. Let Ei be the event that the ith man selects his own hat.
✘ Let i1 < i2 < · · · < ir . Then |Ei1 Ei2 · · · Eir | = (N − r)!,
P (Ei1 · · · Eir ) = Nr · (NN−r)! = r!1 .
P

✓ !
i1 <···<ir

SN N N
(−1)r+1 (−1)r+1 (1/r!).
P P P
✘ P( i=1 Ei ) = P (Ei1 · · · Eir ) =
r=1 i1 <···<ir r=1

The desired probability

TN SN PN
✘ P ( i=1 Eic ) = 1 − P ( i=1 Ei ) = r=0 (−1)r (1/r!).

Examples
✔ Runs. Randomly arrange m black and n white balls in a row.
✘ A run is a consecutive sequence of balls of the same colour.
Find the probability of having r runs in black balls.
Solution. All permutations of m black and n white are equally likely.
✘ Suppose the ith run of black balls has xi balls.
✓ x1 + · · · + xr = m, where xi are positive integers.
✘ White balls are inserted in the gaps, or before, or after black balls.
✓ Let y1 be the number of white balls before the 1st black run.
✓ Let yi be the number of white between (i − 1)th and ith black run.
✓ Let yr+1 be the number of white balls after the r th black run.
✓ y1 + · · · + yr+1 = n, y1 ≥ 0, yn ≥ 0, yi ≥ 1 for i = 2, . . . , r.

16
Examples
✔ Runs. Randomly arrange m black and n white balls in a row.
✘ A run is a consecutive sequence of balls of the same colour.
Find the probability of having r runs in black balls. .
Solution. All permutations of m black and n white are equally likely.
✘ x1 + · · · + xr = m, xi ≥ 1: It has m−1


r−1
solutions.
✘ y1 + · · · + yr+1 = n, y1 ≥ 0, yn ≥ 0, yi ≥ 1, i = 2, . . . , r.
✓ Let z1 = y1 + 1, zr+1 = yr+1 + 1 and zi = yi for i = 2, . . . , r.
z1 + · · · + zr+1 = n + 2, zi ≥ 1: It has n+1


r
solutions.
The probability of having r runs in black balls is m−1

n+1
m+n

r−1 r
/ m .
Similarly, the probability of s runs in white balls is n−1 m+1


m+n

s−1 s
/ m .

Probability as a Continuous Set Function

Monotone Sequence of Events

✔ Definition. Let E1 , E2 , . . . be a sequence of events.
✘ It is monotone increasing if E1 ⊆ E2 ⊆ E3 ⊆ · · · .
S∞
✓ The limit is lim En = n=1 En .
n→∞
✘ It is monotone decreasing if E1 ⊇ E2 ⊇ E2 ⊇ · · · .
T∞
✓ The limit is lim En = n=1 En .
n→∞
✔ Examples.
✘ Let En = {1, 2, . . . , n}.
S∞
✓ Then E1 , E2 , . . . is increasing with lim En = n=1 En = N.
n→∞

✘ Let En = (− n1 , n1 ).
T∞
✓ Then E1 , E2 , . . . is decreasing with lim En = n=1 En = {0}.
n→∞

17
Limit of Events
✔ Let E1 , E2 , . . . be an increasing sequence of events.
c
✘ Let F1 = E1 , Fn = En−1 En . Then F1 , F2 , . . . are mutually exclusive.
Sn S∞ S∞
✓ Note that i=1 Fi = En and i=1 Fi = i=1 Ei = lim En .
n→∞
S∞ P∞
P ( lim En ) = P ( i=1 Fi ) = i=1 P (Fi )
n→∞
n
P
= lim P (Fi ) = lim P (En ).
n→∞ i=1 n→∞

✔ Let E1 , E2 , . . . be a decreasing sequence of events.

✘ Then E1c , E2c , . . . is increasing, and P ( n=1 Enc ) = lim P (Enc ).
S∞
n→∞
T∞
✓ 1 − P ( n=1 En ) = 1 − lim P (En ).
n→∞

✘ Equivalently, P ( lim En ) = lim P (En ).

n→∞ n→∞

Limit of Events
✔ Proposition. Let E1 , E2 , . . . be a sequence of events.
✘ If it is monotone increasing or decreasing, then
✓ P ( lim En ) = lim P (En ).
n→∞ n→∞
This is known to be the continuity of the probability.

✔ Example. Consider a population consisting of individuals able to produce offspring of the

same kind.
✘ Let En be the event that the nth generation is void.
✓ Then E1 , E2 , . . . is a monotone increasing sequence of events.
lim P (En ) = P ( lim En ) = P ( ∞
S
✘
n→∞ n→∞ n=1 En ).

✓ This is the probability that the population even dies out.

18
Probability as a Measure of Belief

Measure of Belief
✔ If n(E) is the number of times E occurs in n repetitions of experiment,
✘ then P (E) = lim n(E)/n.
n→∞
If we believe that a coin is fair,
✘ then P ({H}) = P ({T }) = 21 .
If we believe that a die is fair,
✘ then P ({1}) = P ({2}) = P ({3}) = · · · = P ({6}) = 61 .
✔ It is logical to suppose that measure of the degree of one’s believe should satisfy the
axioms of probability.
✘ For example, in answering Multiple Choices Equations (MCQ),
✓ If one is uncertain about the correct answer, he will choose the answer that he
believes to be mostly likely correct.

Example
✔ Consider a 7-horse race.
✘ One believes that
✓ horse 1 and horse 2 each has 20% chance of winning;
✓ horse 3 and horse 4 each has 15% chance of winning;
✓ horse 5, horse 6 and horse 7 each has 10% chance of winning.
Then according to his belief, the winning probability of
✓ horses 1 or 2 or 3 is 0.2 + 0.2 + 0.15 = 0.55;
✓ horses 1 or 5, 6, 7 is 0.2 + 0.1 + 0.1 + 0.1 = 0.5.
✔ Remark. Suppose subjective probabilities consist with probability axioms.
✘ We deal with idealized person rather than actual person.

19