Special Issue: 25 years of the Mizar Mathematical Library

FORMALIZED MATHEMATICS
Vol. 22, No. 2, Pages 119–123, 2014
DOI: 10.2478/forma-2014-0014 degruyter.com/view/j/forma

Bertrand’s Ballot Theorem1

Karol Pąk
Institute of Informatics
University of Białystok
Sosnowa 64, 15-887 Białystok
Poland

Summary. In this article we formalize the Bertrand’s Ballot Theorem

based on [17]. Suppose that in an election we have two candidates: A that receives
n votes and B that receives k votes, and additionally n ­ k. Then this theorem
states that the probability of the situation where A maintains more votes than
B throughout the counting of the ballots is equal to (n − k)/(n + k).
This theorem is item #30 from the “Formalizing 100 Theorems” list mainta-
ined by Freek Wiedijk at http://www.cs.ru.nl/F.Wiedijk/100/ .

MSC: 60C05 03B35

Keywords: ballot theorem; probability
MML identifier: BALLOT 1, version: 8.1.03 5.23.1210

The notation and terminology used in this paper have been introduced in the
following articles: [24], [1], [14], [15], [18], [4], [5], [10], [21], [6], [12], [3], [11],
[25], [26], [16], [8], [13], [23], and [9].

1. Preliminaries

From now on D, D1 , D2 denote non empty sets, d, d1 , d2 denote finite

0-sequences of D, and n, k, i, j denote natural numbers.
Now we state the propositions:
(1) XFS2FS(dn) = XFS2FS(d)n.
(2) rng d = rng XFS2FS(d).
(3) Let us consider a finite 0-sequence d1 of D1 and a finite 0-sequence d2 of
D2 . If d1 = d2 , then XFS2FS(d1 ) = XFS2FS(d2 ).
1
The paper has been financed by the resources of the Polish National Science Centre granted
by decision no DEC-2012/07/N/ST6/02147.
c 2014 University of Białystok
CC-BY-SA License ver. 3.0 or later
119 ISSN 1426–2630(Print), 1898-9934(Online)

- 10.2478/forma-2014-0014
Downloaded from PubFactory at 08/08/2016 10:53:46PM
via free access
 120 karol pąk

(4) If XFS2FS(d1 ) = XFS2FS(d2 ), then d1 = d2 . Proof: For every i such

that i < len d1 holds d1 (i) = d2 (i) by [2, (13), (11)]. 
(5) Let us consider a finite sequence d of elements of D.
Then XFS2FS(FS2XFS(d)) = d.
(6) Let us consider a finite sequence f and objects x, y. Suppose
(i) rng f ⊆ {x, y}, and
(ii) x 6= y.
Then f −1 ({x}) + f −1 ({y}) = len f .
(7) Let us consider functions f , g. Suppose f is one-to-one. Let us consider
an object x. If x ∈ dom f , then Coim(f · g, f (x)) = Coim(g, x). Proof:
Set f3 = f · g. Coim(f3 , f (x)) ⊆ Coim(g, x) by [6, (11), (12)]. 
(8) Let us consider a real number r and a real-valued finite sequence f . Sup-
pose rng f ⊆ {0, r}. Then f = r · f −1 ({r}). Proof: Define P[natural
P

number] ≡ for every real-valued finite sequence f such that len f = $1 and
rng f ⊆ {0, r} holds f = r · f −1 ({r}). P[0] by [8, (72)]. For every n
P

such that P[n] holds P[n + 1] by [22, (55)], [8, (74)], [25, (70)], [2, (11)].
For every n, P[n] from [2, Sch. 2]. 

2. Properties of Elections

In the sequel A, B denote objects, v denotes an element of {A, B}n+k , and

f , g denote finite sequences.
Let us consider A, n, B, and k. The functor Election(A, n, B, k) yielding a
subset of {A, B}n+k is defined by
(Def. 1) v ∈ it if and only if v −1 ({A}) = n.
Let us note that Election(A, n, B, k) is finite. Now we state the propositions:
(9) Election(A, n, A, 0) = {n 7→ A}. Proof: Election(A, n, A, 0) ⊆ {n 7→ A}
by [19, (29)], [9, (33)], [21, (9)]. 
(10) If k > 0, then Election(A, n, A, k) is empty.
Let us consider A and n. Let k be a non empty natural number. Let us
observe that Election(A, n, A, k) is empty. Now we state the proposition:
(11) Election(A, n, B, k) = Choose(Seg(n+k), n, A, B). Proof: Election(A, n,
B, k) ⊆ Choose(Seg(n + k), n, A, B) by [7, (2)]. 
Let us assume that A 6= B. Now we state the propositions:
(12) v ∈ Election(A, n, B, k) if and only if v −1 ({B}) = k. The theorem is a
consequence of (6).
n+k

(13) Election(A, n, B, k) = n . The theorem is a consequence of (11).

- 10.2478/forma-2014-0014
Downloaded from PubFactory at 08/08/2016 10:53:46PM
via free access
 Bertrand’s ballot theorem 121

3. Properties of Dominated Elections

Let us consider A, n, B, and k. Let v be a finite sequence. We say that v is

an (A, n, B, k)-dominated-election if and only if
(Def. 2) (i) v ∈ Election(A, n, B, k), and
(ii) for every i such that i > 0 holds (vi)−1 ({A}) > (vi)−1 ({B}).
Let us assume that f is an (A, n, B, k)-dominated-election. Now we state
the propositions:
(14) A 6= B.
(15) n > k. The theorem is a consequence of (14) and (12).
Now we state the propositions:
(16) If A 6= B and n > 0, then n 7→ A is an (A, n, B, 0)-dominated-election.
(17) If f is an (A, n, B, k)-dominated-election and i < n−k, then f a (i 7→ B)
is an (A, n, B, (k + i))-dominated-election. The theorem is a consequence
of (14) and (12).
(18) Suppose f is an (A, n, B, k)-dominated-election and g is an (A, i, B,
j)-dominated-election. Then f a g is an (A, (n + i), B, (k + j))-dominated-
election. The theorem is a consequence of (14), (12), and (15).
Let us consider A, n, B, and k. The functor DominatedElection(A, n, B, k)
yielding a subset of Election(A, n, B, k) is defined by
(Def. 3) f ∈ it if and only if f is an (A, n, B, k)-dominated-election.
(19) If A = B or n ¬ k, then DominatedElection(A, n, B, k) is empty. The
theorem is a consequence of (14) and (15).
(20) If n > k and A 6= B, then n 7→ Aa (k 7→ B) ∈ DominatedElection(A, n, B,
k). The theorem is a consequence of (17) and (16).
(21) If A 6= B, then DominatedElection(A, n, B, k) =
DominatedElection(0, n, 1, k). Proof: Set T = [A 7−→ 0, B 7−→ 1]. De-
fine P[object, object] ≡ for every f such that f = $1 holds T · f = $2 .
For every object x such that x ∈ DominatedElection(A, n, B, k) the-
re exists an object y such that y ∈ DominatedElection(0, n, 1, k) and
P[x, y] by [25, (27), (26)], [5, (92)], (7). Consider C being a function from
DominatedElection(A, n, B, k) into DominatedElection(0, n, 1, k) such that
for every object x such that x ∈ DominatedElection(A, n, B, k) holds
P[x, C(x)] from [7, Sch. 1]. DominatedElection(0, n, 1, k) ⊆ rng C by [25,
(27), (26)], [5, (92)], (7). 
(22) Let us consider a finite sequence p of elements of N. Then p is a (0, n, 1,
k)-dominated-election if and only if p is an (n+k)-tuple of {0, 1} and n > 0
and p = k and for every i such that i > 0 holds 2 · (pi) < i. Proof:
P P

If p is a (0, n, 1, k)-dominated-election, then p is an (n + k)-tuple of {0, 1}

- 10.2478/forma-2014-0014
Downloaded from PubFactory at 08/08/2016 10:53:46PM
via free access
122 karol pąk

and n > 0 and p = k and for every i such that i > 0 holds 2 · (pi) < i
P P

by (8), (12), (15), [25, (70)]. 1· p−1 ({1}) = k. p−1 ({1}) + p−1 ({0}) = len p.
1 · (pi)−1 ({1}) = (pi). (pi)−1 ({1}) + (pi)−1 ({0}) = len(pi). 
P

(23) If f is an (A, n, B, k)-dominated-election, then f (1) = A. The theorem

is a consequence of (15).
(24) Let us consider a finite 0-sequence d of N. Then d ∈ Domin0 (n + k, k) if
and only if h0i a XFS2FS(d) ∈ DominatedElection(0, n + 1, 1, k). Proof:
Set X1 = XFS2FS(d). Set Z = h0i. Set Z1 = Z a X1 . Reconsider D = d as a
finite 0-sequence of R. XFS2FS(d) = XFS2FS(D). If d ∈ Domin0 (n+k, k),
then Z1 ∈ DominatedElection(0, n + 1, 1, k) by [15, (20)], (2), [4, (31),
(22)]. Z1 is an (n + 1 + k)-tuple of {0, 1}. For every k such that k ¬ dom d
holds 2 · (dk) ¬ k by [20, (14)], [8, (76)], (1), (3). d is dominated by 0.
P
P
d = k. 
(25) Domin0 (n + k, k) = DominatedElection(0, n + 1, 1, k). Proof: Set D =
Domin0 (n + k, k). Set B = DominatedElection(0, n + 1, 1, k). Set Z = h0i.
Define F[object, object] ≡ for every finite 0-sequence d of N such that
d = $1 holds $2 = Z a XFS2FS(d). For every object x such that x ∈ D
there exists an object y such that y ∈ B and F[x, y]. Consider f being
a function from D into B such that for every object x such that x ∈ D
holds F[x, f (x)] from [7, Sch. 1]. 
(26) Domin0 (n + k, k) = DominatedElection(0, n + 1, 1, k). Proof: Set D =
Domin0 (n + k, k). Set B = DominatedElection(0, n + 1, 1, k). Set Z = h0i.
Define F[object, object] ≡ for every finite 0-sequence d of N such that
d = $1 holds $2 = Z a XFS2FS(d). For every object x such that x ∈ D
there exists an object y such that y ∈ B and F[x, y]. Consider f being
a function from D into B such that for every object x such that x ∈ D
holds F[x, f (x)] from [7, Sch. 1]. 
n−k n+k

(27) If A 6= B and n > k, then DominatedElection(A, n, B, k) = n+k · k .
The theorem is a consequence of (21) and (26).

4. Main Theorem

(28) Bertrand’s Ballot Theorem:

n−k
If A 6= B and n ­ k, then P(DominatedElection(A, n, B, k)) = n+k . The
theorem is a consequence of (13), (19), and (27).

References
[1] Grzegorz Bancerek. Cardinal numbers. Formalized Mathematics, 1(2):377–382, 1990.
[2] Grzegorz Bancerek. The fundamental properties of natural numbers. Formalized Mathe-
matics, 1(1):41–46, 1990.

- 10.2478/forma-2014-0014
Downloaded from PubFactory at 08/08/2016 10:53:46PM
via free access
 Bertrand’s ballot theorem 123

[3] Grzegorz Bancerek. The ordinal numbers. Formalized Mathematics, 1(1):91–96, 1990.
[4] Grzegorz Bancerek and Krzysztof Hryniewiecki. Segments of natural numbers and finite
sequences. Formalized Mathematics, 1(1):107–114, 1990.
[5] Czesław Byliński. Finite sequences and tuples of elements of a non-empty sets. Formalized
Mathematics, 1(3):529–536, 1990.
[6] Czesław Byliński. Functions and their basic properties. Formalized Mathematics, 1(1):
55–65, 1990.
[7] Czesław Byliński. Functions from a set to a set. Formalized Mathematics, 1(1):153–164,
1990.
[8] Czesław Byliński. The sum and product of finite sequences of real numbers. Formalized
Mathematics, 1(4):661–668, 1990.
[9] Czesław Byliński. Some basic properties of sets. Formalized Mathematics, 1(1):47–53,
1990.
[10] Agata Darmochwał. Finite sets. Formalized Mathematics, 1(1):165–167, 1990.
[11] Artur Korniłowicz. On the real valued functions. Formalized Mathematics, 13(1):181–187,
2005.
[12] Rafał Kwiatek. Factorial and Newton coefficients. Formalized Mathematics, 1(5):887–890,
1990.
[13] Beata Padlewska. Families of sets. Formalized Mathematics, 1(1):147–152, 1990.
[14] Karol Pąk. Cardinal numbers and finite sets. Formalized Mathematics, 13(3):399–406,
2005.
[15] Karol Pąk. The Catalan numbers. Part II. Formalized Mathematics, 14(4):153–159, 2006.
doi:10.2478/v10037-006-0019-7.
[16] Jan Popiołek. Introduction to probability. Formalized Mathematics, 1(4):755–760, 1990.
[17] M. Renault. Four proofs of the ballot theorem. Mathematics Magazine, 80(5):345–352,
December 2007.
[18] Andrzej Trybulec. Domains and their Cartesian products. Formalized Mathematics, 1(1):
115–122, 1990.
[19] Andrzej Trybulec. Enumerated sets. Formalized Mathematics, 1(1):25–34, 1990.
[20] Andrzej Trybulec. On the decomposition of finite sequences. Formalized Mathematics, 5
(3):317–322, 1996.
[21] Andrzej Trybulec. Binary operations applied to functions. Formalized Mathematics, 1
(2):329–334, 1990.
[22] Wojciech A. Trybulec. Non-contiguous substrings and one-to-one finite sequences. For-
malized Mathematics, 1(3):569–573, 1990.
[23] Zinaida Trybulec. Properties of subsets. Formalized Mathematics, 1(1):67–71, 1990.
[24] Tetsuya Tsunetou, Grzegorz Bancerek, and Yatsuka Nakamura. Zero-based finite sequ-
ences. Formalized Mathematics, 9(4):825–829, 2001.
[25] Edmund Woronowicz. Relations and their basic properties. Formalized Mathematics, 1
(1):73–83, 1990.
[26] Edmund Woronowicz. Relations defined on sets. Formalized Mathematics, 1(1):181–186,
1990.

Received June 13, 2014

- 10.2478/forma-2014-0014
Downloaded from PubFactory at 08/08/2016 10:53:46PM
via free access