Mathematical Induction Solutions

Mathematical Induction Solutions

Problems
, so the base case is finished.
1.(a) Base Step: 1 = 1(1+1)
2
Induction Step: Suppose the result holds for some positive integer n. We need to show
the result holds for n + 1. We have:
1 + 2 + . . . + n + 1 = (1 + 2 + . . . + n) + (n + 1)
n(n + 1)
+ (n + 1)
since the result holds for n
=
2
(n + 1)(n + 2)
=
2
hence the result holds for n + 1 as well, and the induction step is complete.
1.(b) Base Step: 12 = 1(1+1)(21+1)
, so the base case is finished.
6
Induction Step: Suppose the result holds for some positive integer n. We need to show
the result holds for n + 1. We have:
12 + 22 + . . . + (n + 1)2 = (12 + 22 + . . . + n2 ) + (n + 1)2
n(n + 1)(2n + 1)
=
+ (n + 1)2
since the result holds for n
6
n+1
n+1
(n(2n + 1) + 6(n + 1)) =
(2n2 + 7n + 6)
=
6
6
(n + 1)(n + 2)(2(n + 1) + 1)
=
6
hence the result holds for n + 1 as well, and the induction step is complete.
1
1
=1
, so the base case is finished.
1.(c) Base Step:
12
1+1
Induction Step: Suppose the result holds for some positive integer n. We need to show
the result holds for n + 1. We have:
1
1
1
1
1
1
1
+
+ ... +
=(
+
+ ... +
)+
12 23
(n + 1) (n + 2)
12 23
n (n + 1)
(n + 1) (n + 2)
1
1
= (1
)+
since the result holds for n
n+1
(n + 1) (n + 2)
n+21
1
=1
=1
(n + 1) (n + 2)
n+2
hence the result holds for n + 2 as well, and the induction step is complete.

Mathematical Induction Solutions

2. The first few terms in the sequence are 1, 3, 7, 15, 31, . . .. We see that adding 1 to each
of these gives us just powers of 2. We claim that the nth term in the sequence has the
form 2n 1. Let us prove this claim using induction.
Base Step: The first term is 1, while 21 1 = 1. So the base case is done.
Induction Step: Assume the result is true for some positive integer n, so that an =
2n 1. For n + 1, we have:
an+1 = 2an + 1 = 2(2n 1) + 1 = 2n+1 2 + 1 = 2n+1 1
so the induction step is complete.
We have thus managed to prove that an = 2n 1, so a300 = 2300 1.
3.(a) Base Step: 51 = 5 which clearly ends in 5, so the base step is done.
Induction Step: Assume the result is true for some positive integer n, so that 5n ends
in 5. Another way to write this is 5n = 10x + 5 for some ineteger n. We want to show
the result for n + 1. We have:
5n+1 = 5 5n = 5 (10x + 5) = 50x + 25 = 50x + 20 + 5 = 10(5x + 2) + 5
which ends in 5, so the induction step is finished.
3.(b) Base Step: (for n = 2 ) 52 = 25 which clearly ends in 25, so the base step is done.
Induction Step: Assume the result is true for some positive integer n 2 (note the
extra condition here, since we started with n = 2 ), so that 5n ends in 25. Another way
to write this is 5n = 100x + 25 for some ineteger n. We want to show the result for
n + 1. We have:
5n+1 = 5 5n = 5 (100x + 25) = 500x + 125 = 500x + 100 + 25 = 100(5x + 1) + 25
which ends in 25, so the induction step is finished.
4. Base Step: (for n = 4 ) 4! = 24, while 24 = 16; thus n! > 2n for n = 4, and the base
step is finished.
Induction Step: Assume the result is true for some positive integer n 4, so that
n! > 2n . For n + 1 we have:
(n + 1)! = (n + 1) n! (n + 1) 2n 2 2n = 2n+1
where the first inequaltiy holds since n! 2n , and the second holds since n + 1 2.
(This is true since n 4, and hence n+1 5 2). We have shown that (n+1)! 2n+1 ,
so the result holds for n + 1, and the induction step is done.
5. This is similar to problem 3.
Base Step: 81 1 = 7, which is clearly divisible by 7, so the base step is complete.
Induction Step: Assume the result is true for some positive integer n, so that 8n 1 is
2

Mathematical Induction Solutions

divisible by 7. Another way to write this is 8n = 7k + 1 for some ineteger k. We want
to show the result for k + 1. We have:
8n+1 1 = 8 8n 1 = 8 (7k + 1) 1 = 56k + 8 = 56k + 7 = 7(8k + 1)
which is divisible by 7, so the induction step is done.
6. Base Step: Clearly (1 + x)1 1 + x, so the base step is done.
Induction Step: Assume the result is true for some positive integer n, so that (1+x)n
1 + nx. We need to show the result for n + 1. We have:
(1 + x)n+1 = (1 + x)(1 + x)n
(1 + x)(1 + nx) since the result holds for n and 1 + x 0, since x 1
= 1 + (n + 1)x + nx2 1 + (n + 1)x since x2 0
therefore (1 + x)n+1 1 + (n + 1)x, so the induction step is complete.

Some Harder Problems

7. Base Step: The result holds for n = 1, because:
1
1
1
1
= < =
21
2
21+1
3
Induction Step: Assume the result is true for some positive integer n, so that:
1
1 3 5 . . . (2n 1)

2 4 6 . . . 2n
2n + 1
The result for n + 1 is similar, except we need to
multiply the left side of the above
2n + 1
2n + 1
equation by
, while the right side by p
. So as long as we show
2n + 2
2(n + 1) + 1
that we multiply the left side by a smaller number, than the right side, the result will
hold. Thus we just need to prove:

2n + 1
2n + 1
p
2n + 2
2(n + 1) + 1
Squaring both sides we get:
(2n + 1)2
2n + 1

(2n + 1)(2n + 3) (2n + 2)2

2
(2n + 2)
2n + 3
4n2 + 8n + 3 4n2 + 8n + 4
3

Mathematical Induction Solutions

which is clearly true. Therefore the result holds for n + 1:

1
1 3 5 . . . (2n 1) 2n + 1
1
2n + 1
p
=p

2 4 6 . . . 2n
2n + 2
2n + 1
2(n + 1) + 1
2(n + 1) + 1
and we are done.
8. We look at the first few terms of the sequence and try to spot a pattern. Have a1 =
1, a2 = 2, a3 = 4, a4 = 8, . . .. This suggests that an = 2n1 . Let us prove this using
strong induction.
Base Step: For n = 1 and n = 2 the result holds: a1 = 1 = 211 and a2 = 1 = 221 .
Note: In this base step, we considered two cases, and not one as usual. We need to
do it for this problem since the n + 1-st term depends on two previous terms, and not
just one.
Induction Step: Let n 2 be a positive integer, and assume the result holds for all
positive integers k n. To prove the result holds for n + 1, we use the definition of
the sequence:
an+1 = an + 2an = 2n1 + 2 2n2 = 2n = 2(n+1)1
where the second equality holds since the result holds for n and n 1, so that an =
2n1 , an1 = 2n2 , by the induction assumption. We have thus shown the result holds
for n + 1, and we are done.
9. We will call a positive integer n good if Bob can pay for an item worth n dollars with
two- and five- dollar bills without change. Its clear that if n is good, then n + 2 must
be good, since Bob just pays with another two-dollar bill. So for a particular positive
integer n to prove that is good, it would suffice to know that n 2 is good and not
just that n 1 is good, so we will again use strong induction.
Base Step: n = 4 is good: Bob can pay using two two-dollar bills.
n = 5 is good: Bob can pay using one five-dollar bill.
Induction Step: Let n be a positive integer, n 5 and suppose for all integers k
satisfying 4 k n, k is good. We need to show that n + 1 is good.
Since n 5, then n 1 4 so k = n 1 satisfies 4 k n and hence n 1 is
good. But then n + 1 is good (remember Bob just uses an extra two-dollar bill), and
therefore n + 1 is good as well, so the induction step is finished.
Note: Make sure you understand the logic in the above argument well, and if not,
re-read the solution. The induction step itself is quite simple, but the logic for setting
it up is more tricky than in the previous problems.
1
10. Base Step: For n = 1 we have 3(2 ) 1 = 8 which is divisible by 23 = 21+2 but not by
24 = 21+3 . So the base step is done.
Induction Step: Assume the result is true for some positive integer n. Then we have
n
n
3(2 ) 1 is divisible by 2n+2 but not 2n+3 . This is the same as saying that 3(2 ) 1 =

Mathematical Induction Solutions

2n+2 k, where k is an odd number.
For n + 1 we have:
n+1
n
3(2 ) 1 = [3(2 ) ]2 1
n
n
= (3(2 ) 1)(3(2 ) + 1)
n
n
= (3(2 ) 1)(3(2 ) 1 + 2)
= (2n+2 k)(2n+2 k + 2)
= 2n+2 k 2 (2n+1 k + 1)
= 2n+3 (k (2n+1 k + 1))

Since n 1, it follows that (2n+1 k + 1) and hence k (2n+1 k + 1) is odd. Therefore

n+1
from the above derivation we obtain that 3(2 ) 1 is divisible by 2n+3 = 2(n+1)+2
but not 2n+4 = 2(n+1)+2 , and the induction step is complete.
11. With 1 line there are 2 regions. If we draw another line, we add 2 more regions. If we
draw a third line, we add 3 more regions. In general, if we have n lines already drawn,
and we draw another line, then new line will intersect each of the other lines exactly
once, and thus will be split up into n + 1 disjoint segments or rays. For each of these
segments or rays, we can mark off the region on the paper when only n regions were
drawn, which contains this segment or ray after it is drawn. This region will be split
up in exactly two regions, thus increasing the count of the regions by 1. Therefore each
disjoint segment or ray creates a distinct new region, and in total n + 1 new regions
are created.
Using the above argument, we can prove by induction that the number of regions
created by n lines is:
n(n + 1)
1 + 1 + 2 + ... + n = 1 +
2
Base Step: For n = 1 we have 2 = 1 + 1 regions, so the result holds.
Induction Step: Assume the result is true for some positive integer n, so that the
number of regions when n lines are drawn is:
1 + 1 + 2 + ... + n
When we draw another line, we get n + 1 lines, while the number of regions increases
by n + 1 (by the argument stated above), hence the number of regions is:
(1 + 1 + 2 + . . . + n) + (n + 1) = 1 + 1 + 2 + . . . + n + 1
so the result is true for n + 1, and the induction step is finished.
12. We will use strong induction.
1
Base Step: For n = 1, x + is an integer by the problem condition.
x
5

Mathematical Induction Solutions

For n = 2, we have:
x2 +

1
1
1
1
= (x + )2 2 x = (x + )2 2
2
x
x
x
x

which is an integer since x + x1 and 2 are both integers. This completes the base step.
Induction Step: Let n 2 be an integer, and assume the result holds for all positive
1
integers k, k n. In particular, this means that xn + n is equal to some integer a,
x
1
n1
and x
+ n1 is equal to some integer b. For n + 1 we have:
x
xn+1 +

1
xn+1

1
1
1
)(x + ) (xn1 + n1 )
n
x
x
x
1
= a(x + ) b
x

= (xn +

1
which must be an integer, since a, b are integers, and by the problem condition, x +
x
is also an integer. Thus the result is true for n + 1 as well, and we are done.

Hard Problems
13. Notice that for (a) and (b) m and n have greatest common divisor equal to 1, therefore
on each turn a player can always make a move of replacing the number k with its
divisor l strictly less than k, as long as l > 1, or as long as l = 1 and 1 has not yet
appeared on the board.
Instead of dealing with the actual numbers we will deal with the number of prime
factors they have. Then, the game becomes equivalent to the following. Two numbers
m and n are written on the board. On each turn a player can select a number k greater
than 0 and replace it with any positive integer less than k, or replace it with 0, as long
as 0 is not already written on the board. A player who cannot make a move loses.
It immediately follows that m = 0, n = 1 is a losing position. Therefore, m = 0, n 2
is a winning position (since a player replaces n with 1 and wins). Furtheremore,
m = 1, n 1 is a winning position (since a player replaces n with 0 and wins). Hence
m = 2, n = 2 is a losing position; m = 2, n 3 is a winning position; m = 3, n = 3 is
a losing position, m = 3, n 4 is a winning position. By induction it follows that for
k 2, m = k, n = k is a losing position, while m = k, n k + 1 is a winning position.
(a) We are in the case of m = 40, n = 51 41 in the transformed game, thus this is
a winning position and Alphonse wins.
(b) This case is different, since now m and n have more than one divisor in common.
We will deal with the original game and not make any transformations. Note that m
6

Mathematical Induction Solutions

and n are both powers of 2, so throughout the whole game only powers of 2 can appear
on the board.
We first note that the player who first writes down a number less than or equal to
2 loses. This is because if they write down 1, then 2 has not yet been written; the
opponent on the next turn replaces the other number with 2 wins. (Note that this
move is legal since at the start m > 2, n > 2 so at the time that 1 is written, the other
number on the board must be greater than 2). If they write down 2, then 1 has not
yet been written; the opponent on the next turn replaces the other number with 1 and
wins.
Similarly, the player who first writes down a number less than or equal to 8 loses. This
is because if they write down 4, the other player writes 8 thus forcing the original
player to write down a number less than or equal to 2 (note they cannot replace 8
with 4 since 4 has already appeared on the board). Similarly, if they write down 8, the
other player writes down 4 and wins.
By induction it follows that if m, n > 22k1 then the player who first writes down a
number less than or equal to 22k1 loses for every positive integer k. Thus for the case
m = 240 , n = 253 , the player to first write down a number less than or equal to 239
loses. Therefore on his first turn, Alphonse replaces 253 with 241 and wins because on
her turn, Beryl is faced with 240 and 241 on the board and has to write down a number
less than or equal to 239 .
Note: you should always consider induction as a possible method for proving that a
particular strategy in a game is a winning one because often the objective in games
is to reduce the current state of the game to a smaller state, which is exactly how
induction works.
14. We will prove that np gives the remainder of n when divided by p. Since p is prime,
np
gives the
if we manage to prove this, then we get the result, since then np1 =
n
n
= 1 when divided by p. As usual, we use induction. This time, we
remainder of
n
need to only prove the result for all integers n between 1 and p 1, inclusive, since for
any positive integer n, we have np gives the same remainder when divided by p, as k p ,
where k is the remainder that n itself gives, when divided by p. (This is a useful fact
in number theory).
Base Step: n = 1. Then 1p = 1, which gives remainder 1 when divided by p. Thus the
base step is done.
Induction Step: Let n be a positive integer, with 1 n p 2, such that the result
holds for n, i.e. np gives remainder n when divided by p. Another way to write this is

Mathematical Induction Solutions

np = kp + n for some integer k. By the binomial theorem, for n + 1 we have:
 
 


p
p p1
p p2
p
p
(n + 1) = n +
n
+
n
+ ...
n+1
1
2
p1
For i = 1, . . . , p 1, the binomial coefficient:
 
p
p(p 1) . . . (p i + 1)
=
i
i!
must be divisible by p, since the numerator is divisible by p, while the denominator is
not (its a product of integers all of whom are less than p and hence cannot be divisible
by p, since it is prime). This means that in the binomial expansion of (n + 1)p above,
all the terms involving binomial coefficients are divisible by p, and hence the remainder
that (n + 1)p gives when divided by p is the same as np + 1 does:
 
 


p p1
p p2
p
p
p
(n + 1) = n +
n
+
n
+ ...
n +1
1
2
p1
{z
}
|
divisible by p

= n + lp + 1
for some integer l. Replacing np with kp + n we get:
(n + 1)p = kp + n + lp + 1 = p(k + l) + (n + 1)
which means that (n + 1)p gives the remainder of (n + 1) when divided by p, and the
induction step is complete.
15. Base Step: n = 4. We have:
x1
x2
x3
x4
+
+
+
x4 + x2 x1 + x3 x2 + x4 x3 + x1
r
x1 + x3 x2 + x4
x1 + x 3 x2 + x 4
=
+
2

=2
x2 + x4 x1 + x3
x2 + x 4 x1 + x 3
using the AM-GM inequality. This completes the base step.
Induction Step: Assume the result holds for some positive integer n; we wish to prove
it for n + 1. Let x1 , x2 , . . . , xn+1 be arbitrary positive real numbers; we wish to show:
A=

x1
x2
xn+1
+
+ ... +
2
xn+1 + x2 x1 + x3
x n + x1

In the above expression, we can without loss of generality assume that xn+1 is the
smallest among x1 , x2 , . . . , xn+1 . This will be used later.
8

Mathematical Induction Solutions

By the induction assumption, we know that:
B=

x1
x2
xn
+
+ ... +
2
xn + x2 x 1 + x3
xn1 + x1

If we can show that A B, then it would immediately follow that A 2 and we

would be done. Let us try to show A B by looking at the difference between these
two numbers:
x1
xn
xn+1
x1
xn
+
+
(
+
)
xn+1 + x2 xn1 + xn+1 xn + x1
xn + x2 xn1 + x1
x1
x1
xn
xn
xn+1
=(

)+(

)+
xn+1 + x2 xn + x2
xn1 + xn+1 xn1 + x1
xn + x1

AB =

The first bracket in the above expression is non-negative since xn+1 + x2 xn + x2 ,

which is true because xn+1 xn (recall that we assumed earlier than xn+1 xi for
all i = 1, 2, . . . n). The second bracket is non-negative since xn1 + xn+1 xn1 + x1 ,
which is true because xn+1 x1 . And the last term is positive since xn+1 , xn , x1 are
all positive. Thus the whole expression is non-negative, and A B 0.
Therefore A B. Since B 2, it follows that A 2, which means that the result is
true for n + 1 and the induction step is done.
Note: The assumption that xn+1 was the smallest among the terms is key here. How
does one come up with such an assumption? Well, we could write out the difference
A B regardless of any assumptions. Then we see, that this would be non-negative
if we can get xn+1 xn and xn+1 xn1 . So as long as in the sequence of numbers
x1 , x2 , . . . , xn+1 , x1 , x2 we can find one that is less than or equal to the previous two,
we are in business. This is of course always possible.