Let us discuss an interesting combinatory problem. Here is the question.
A local high school math club has 12 students in it. Each week, 6 of the students go on a field trip.
(a) Jeffrey, a student in the math club, has been on a trip with each other student in the math club. Determine the minimum number of trips that Jeffrey could have gone on.
(b) If each pair of students have been on at least one field trip together, determine the minimum number of field trips that could have happened.
Any clue on how to work with this problem?
Firstly, this is a combinatorial problem. The questions are looking for a number — a naive way is to make a schedule and count for the number.
However if we arbitrarily make a schedule, even in the best effort, can we guarantee that the minimum number of trips is reached in our scheduling (both for Jeffery, and for the total number of trips). No. The minimum can not be reached that easily.
A better way is to show there is a minimum n; for less than n trips, the restricting condition becomes impossible. Take the case for Jeffery (referring to question (a)). Suppose Jeffery has been in two trips. In the first trip, he was with five students A1, A2, A3, A4, A5; in his second trip, he was with B1, B2, B3, B4, B5 (assume only Jeffery took both these trips). Concludingly, he has been with ten other students. Adding himself, we counted 11 students. But recall there are 12 students, so there is a C who has not been in any of these two Jefferey’s trips. Therefore, 3 is the minimum number of trips if, Jeffery has been together with any of the other 11 students for at least one trip – 3rd trip is necessary for Jeffery to be with C together
The question (b) has a more involved nature. First point to consider: in what way can count to ensure any two students have been together in at least one field trip? Arguing for the minimum, we know too less trips will not generate an arrangement ensuring the condition: any two students are together at least once; however, if we do reach the minimum number of trips, would that condition be satisfied? — that is the second point to consider. If we want to guarantee the condition satisfied, then the obvious way is to actually make a schedule so as to check that condition — in this example, we need to show that in some n trips, any two of the 12 students have been in a field trip together. So the clues are provided for your — dear reader; would you like to carry it further?
We leave to our readers to complete this question. ** Reference solutions will be posted soon. **
Simple Diophantus Equations
Take a look at the following question.
Determine all integers a for which \frac{a}{1011-a} is an even integer.
Hint: Let \frac{a}{1011-a} = 2d . Solve for a in terms of d.
(In the usual sense — if seeking for real roots, Diaphantus equations has infinitely many roots; but is has finite many roots if only the integer roots are sought.)
Suppose the expression \frac{a}{1011-a} gives an even integer 2d (d is an integer). This leads to
a = 2022 d – 2ad, which is (2d+1)a + (2) \cdot 1011d so a = (2) \frac{1011 d}{2d+1} = (2) \frac{(3)(337)d}{2d+1} -- -- (*)
Note a is integer and d is integer. By inspecting (*): the basic clue is that any prime factor of (2d+1) (in denominator) must be cancelled by the factors in the numerator — so that the quotient a is an integer. Note d and 2d +1 are co-prime; and numbers 3 and 337 are both prime numbers.
Therefore 2d +1 must take one of the following values: \pm1, \pm3, \pm 337, \pm1001 .
Here comes the last step to find the value of a. We leave it to the reader to complete.
By the way, can you see that in all solutions, a must be positive (i.e. a >0)? Yes, that is indeed true; otherwise if a <0, then the quotient will not be an integer — in fact, the absolute value of quotient will be between 0 and 1 (please figure out why).
Suppose there are three sequences {an}, {bn}, and {cn}, where n=1, 2, 3 … . These sequences are defined recursively, and with inter-dependence; for example, the definition of bn depends not only on bn-1, but also on an-1 or cn-1. Suppose we are to work out some inequality relations among an, bn, and cn. In what way can we make it?
Here is one such question, as follows.
Let p, q be any two distinct positive real numbers such that pq = 5.
Define A_0 = -1 - p - q, B_0 = p+q+pq, C_0 = - pq <0.
For n \geq 1 , define
Show that 1+B_n > (-C_{n-1} +1)^2 >0 for any n \geq 1 .
Let us first calculate B1. Using the recursive definition, we have that B_1 = (B_0)^2 - 2 A_0 C_0 \\ = (p+q+pq)^2 - 2 (-1-p-q) (-pq) \\ = p^2 + q^2 + p^2 q^2 >=0. Note it is obvious that B1 is non-negative. Continuing in this way, any Bn shall be non-negative.
Let us set the next immediate goal as 1+B_1 > (-C_0 +1)^2 . Bringing the value of C_0 = -pq , and noting that B_1 = p^2 + q^2 + p^2 q^2 , the goal of proof is equivalent to 1 + p^2 + q^2 + p^2 q^2 > (1-pq)^2; which shall be obvious as p^2 + q^2 > 0 > -2pq .
We list the values of {A0, B0, C0; A1, B1, C1}, in terms of p, q, as follows: (previously, we show how B1 can be calculated; the read can try to find A1, C1, similarly)
A_0 = -1-p-q, B_0 = p+q+pq, C_0 = -pq, \\ A_1 = -1 - p^2 -q^2, B_1 = p^2 + q^2+p^2 q^2, C_1 = -p^2 q^2 .
The quick-minded reader must have found that, when advancing the index by 1, the following schema holds true (where p^*, q^* are two distinct positive real numbers), A_{n-1} = -1-p^*-q^*, C_{n-1} = -p^* q^*, \\B_{n-1} = p^*+q^*+p^*q^*, \\ A_{n} = -1 - (p^*)^2 -(q^*)^2, C_{n} = -{p^*}^2 {q^*}^2, \\ B_{n} = {p^*}^2+{q^*}^2+{p^*}^2 {q^*}^2.
So, a straightforward generalization for the proof of 1+B_1 > (-C_0 +1)^2 , leads to the proof of 1+B_n > (-C_{n-1}+1)^2 .
We have two remarks in sequel.
The following question comes from 2024 CSMC (Canadian Senior Math Contest) – Part B, Question 2. (There are three sub-questions, a, b, and c. We present only a. and c.)
a. The quadratic equation x2-2x-1= 0 has solutions x= r and x = s. Determine integers b and c for which quadratic equation x2 + bx + c = 0 has solutions x = 2r + s and x = r + 2s.
c. Suppose that A1=-6, B1 = 10 and C1 = -5. For each positive integer n \geq 2 , let \\ A_n = 2 B_{n-1} – (A_{n-1})^2, \\ B_n = (B_{n-1})^2 – 2 A_{n-1} C_{n-1}, \\ C_n = -(C_{n-1})^2 .
Prove that the polynomial f_{100}(x) = x^3 + A_{100} x^2 + B_{100} x + C_{100} has three distinct positive real roots.
Now we look into how to solve the question above.
Analysis for a.
It shall hold x2 + px + q = (x-r) (x-s) if r, s are two roots of the equation. By expanding the right hand side, we obtain x2 – (r+s)x + rs. Now comparing it to x2 + px + q to find [this is indeed Viete’s Theorem]
r+s = -p, (rs) = q ~---~ -- (*)
Using (*) for the question in 2a), it holds that r+s = -p =2, and (rs) =q = -1; meanwhile we have -b = (2r+s) + (r+2s) = 3(r+s) etc. from which the value of b can be calculated.
Analysis for c.
We intend to show that for any n, equation f_n (x) = x^3 +A_n x^2 +B_n x + C_n has three distinct roots. (This is actually stronger than what the question is asked; yet if this is indeed true, then a clear path can be paved leading all the way to f100 (x). )
If one of the three roots (call it x1) is known (preferably an integer), then we have a linear factor (x – x_1) (by Factor Theorem), and the co-factor (quadratic) can be worked out with polynomial division. For the said quadratic co-factor, we argue it has real distinct roots by checking that the discriminant is positive, and be ensured neither root of the quadratic factor equals x_1 .
Note we find x=1 is a root for fn(x) (so let x1 =1.) And f1(x), f2(x), … each has three distinct roots. [Details to be worked out]
Next step: The readers are trusted to work out the solution to a. (analysis for a. was already given). For question c., in addition to the analysis given above, there are some gaps need to be filled. Please attempt to work them out by yourself, yet if you get stuck or want to check whether you have done it right, check out the next post. Using Factor Theorem (cont.) – selected details for a CSMC question
- For year 2026, the math contest CSMC/CIMC (Canadian Senior and Intermediate Math Contest) will be held on Wednesday, November 18th. CSMC is recommended to students in Grade 11 and 12 to participate, while CIMC is for students in Grade 9 and 10.
- If you are new to the math contests in Canada, please read this post.
- For year 2026, COMC (Canadian Open Math Challenge) will be held on October 29th. Recommended for students in grade 7-12 who love the challenge of math. — Registration will be opening soon.
As shown in the figure below: (please scroll down a bit to see the figure) AB’ is one diameter of the circle, and a blue curve from A to B is supposed to bisect the area of the circle. We see from the figure below: 
Length of the blue curve AB is greater than: AE + EB = AE + EB’
which is certainly longer than AB’; AB’ is the diameter.
The blue curve is obvious the focus. (Note when blue curve AB is mentioned, we mean the “curved one”: it curves around, not going straight from A to B.) Besides, point E is where the blue curve intersects with CD; and CD is one diameter.
So we have proved the claim that any curve bisecting the circular area got to be longer than the diameter.
Stay with us for one more minute. Let the construction-proof process be revealed to you, as follows.
Referring back to the figure. Let us start from the circle and the blue curve only (imagine all other lines and points disappear; now we draw them step by step). Connect the two endpoints A, B of the blue curve by a line segment AB. Then draw the diameter CD // AB (i.e. line CD is parallel to line AB).
Take point O (the midpoint of CD), and then passing A and O, let another diameter AB’ be drawn.
For completing the proof , two arguments are required:
(1) The blue curve has at least one intersection point with diameter CD (thinking it: if the blue curve resides at only one side of CD, then that curve cannot divide the circular area evenly into two parts with equal area; therefore, any curves that bisects the circular area must intersects diameter CD). Now suppose the intersection point is E.
(2) B and B’ are symmetric to the diameter CD therefore EB = EB’ (trying justifying it using the property of circles and parallel lines).
The rest of the proof is straightforward (and intuitive).
(Revised in August 20, 2026)
Have you heard about the Canadian Math Contests? It is certainly for those talented in math and logic to show their ability, and a good opportunities for those who love the challenges.
For a student to participate in a contest, talk to the school one attends, or talk to us (Jonah’s Math Corner). Be sure to have a current enrollment in a secondary school (junior, high school) — or sometimes elementary school, located in Canada.
Check your grade too — if it says grade 9, then all students in grade 10 and higher cannot write that contest. A students in lower grade are allowed to participate in contests for higher grade (only if he or she feels confident), but not the other way round.
If you are looking at very selective competitions, then COMC (Canadian Open math challenge) is the contest to participate (typically held in October). The very best performed students can earn an entry to represent Canada for the IMO (International Math Olympiad) through further selective process.
Students in grade 9/10/11 may look for CIMC (Canadian Intermediate Contest) and CSMC (Canadian Senior Math Contest) — here Senior means “Senior High School” — Both are held in November on the same time, but no one is allowed to participate in both (just choose one of them).
Keep reading for information about different Contest series, detailed as follows:
1) Junior math contests (have to be under grade 9 to participate): the most popular is the Gauss contest for grades 7 & 8, organized by U. of Waterloo. Gauss is very popular that some school teachers enrolled all of their students to participate.
There are some grade 9 contest – see the mid-level and high-school level contests, as the challenge is already close to high levels.
If you live in Alberta, then both U. of Calgary and U. of Alberta have set up challenges for Junior students.
2) Mid-level and High-school level Contests — the main stage for the middle school math challenge. University of Waterloo has organized 3 series respectively for grade 9, 10 and 11. The contests questions appears typically in multiple-choice format.
Series I: PCF contest: Pascal for grade 9, Cayley for grade 10 and Fermat – for grade 11.
Questions in PCF appear in multiple-choice format.
Series II: FGH contest: Fryer – for grade 9, Galois – for grade 10 and Hypatia – for grade 11.
No multiple-choice question in FGH series. All answers have to be worked out by participants. In any of these series, there will be two types of questions, one will ask participants for answers only, and the other type will ask students to write full solutions. Each contests is 75 minutes long.
Series III: CIMC – Canadian Intermediate Math Contests (for grade 9/10) and CSMC – Canadian Senior Math Contests (for grade 11/12). The format is similar to FGH, however, the level of challenge is in general higher, and students are provided with 2 hours time to answer.
4) Selective Contests for those best talents in math: Canadian Math Society have a challenge series for those who want to participate in Math Olympiad. The first step is the COMC (Canadian Open Math Challenge). It is selective, so only COMC is open to students; all following contests in the series, like CMO (Canadian Math Olympiad) and IMO (International Math Olymiad), are by-invitation ONLY.
5) Popular Math Contests: Math Kangaroo, started in Europe and becoming popular across world, is organized by Math Kangaroo company. Their are 6 age groups and each group bracket two grades, like grades 1-2, grades 3-4, .. .. According to the organizer’s claim, the purpose of the contest is to challenge students in a playful setting; instead of academic, it aims at developing mental powers and flexibility.
5) Other Contests: besides those mentioned, we occasionally will register students in Regional Math League of Canada or US – some contests are online only – save students the cost and time to travel.
Jonah’s math corner can serve as a site to supervise students to write the math contests. If you have any question, just talk to us.
Questions chosen from previous Gauss contests
Gauss contests are organized by the Centre of Education for
Math and Computing, University of Waterloo
Problem 1
In the addition shown, P and Q each represent single digits, and the sum is 1PP7. What is P + Q?
(A) 9 (B) 12 (C) 14 (D) 15 (E) 13
Problem 2
In the right-angled triangle PQR, we have that PQ = QR. The three segments QS, TU and VW are perpendicular to PR, and the segments ST and UV are perpendicular to QR, as shown. What fraction of triangle PQR is shaded?
(A) 3 ⁄ 16 (B) 3 ⁄ 8 (C) 5 ⁄ 16 (D) 5 ⁄ 32 (E) 7 ⁄ 32
Problem 3
A box contains a total of 400 tickets that come in five colors: blue, green, red, yellow, and orange. The ratio of blue to green to red tickets is 1 : 2 : 4. The ratio of green to yellow to orange tickets is 1 : 3 : 6. What is the smallest number of tickets that must be drawn to ensure that at least 50 tickets of the same colour have been selected?
(A) 50 (B) 246 (C) 148 (D) 196 (E) 115
Problem 4
Greg, Charlize, and Azarah run at different but constant speeds. Each pair ran a race on a track that measured 100 m from start to finish. In the first race, when Azarah crossed the finish line, Charlize was 20 m behind. In the second race, when Charlize crossed the finish line, Greg was 10 m behind. In the third race, when Azarah crossed the finish line, how many metres was Greg behind?
(A) 20 (B) 25 (C) 28 (D) 32 (E) 40
Problem 5
In right-angled, isosceles triangle FGH, segment FH = √̅8. Arc FH is part of the circumference of a circle with centre G and radius GH. The area of the shaded region is
(A) π – 2; (B) 4 π – 2 (C) 4 π – (1 ⁄ 2) √̅8 ; (D) 4 π – 4 (E) π – √̅8
This posts follows a previous post. For the convenience of readers, the question is repeated:
Suppose that A1=-6, B1 = 10 and C1 = -5. For each positive integer n \geq 2 , let \\ A_n = 2 B_{n-1} – (A_{n-1})^2, \\ B_n = (B_{n-1})^2 – 2 A_{n-1} C_{n-1}, \\ C_n = -(C_{n-1})^2 .
Prove that the polynomial f_{100}(x) = x^3 + A_{100} x^2 + B_{100} x + C_{100} has three distinct positive real roots.Selected Details for the solution of c.
Now we start to work out the problem just-shown.
We find that 1+A_1 + B_1 + C_1 = 0 , then f_1 (x) = x^3 + A_1 x^2 + B_1 x + C_1 = (x-1) (x^2 + (A_1+1) x - C_1) ; (using polynomial division \frac{x^3 + A_1 x^2 + B_1 x + C_1}{x -1} to find the quadratic cofactor). By bringing A1=-6, C1= -5 into the factorized form of f_1(x) :
f_1 (x) = (x-1) (x^2 -5 x +5).
Obviously, x^2 – 5x +5 = 0 has two distinct roots [as the discriminant is (-5)2 – 4(5) = 5 >0; and note for this quadratic, none of the roots is 1]. Thus f_1(x) has three distinct roots.
In general, if 1 + A_n + B_n + C_n = 0 (this is indeed the case), then
f_n (x) = x^3 + A_n x^2 + B_n x + C_n \\ = (x-1) (x^2 + (A_n+1) x – C_n).
To show that 1 + A_n + B_n + C_n = 0 , we proceed as follows: 1 + A_n + B_n + C_n \\ = 1 + (2B_{n-1} - A^2_{n-1}) + (B^2_{n-1} - 2 A_{n-1} C_{n-1}) + (- C^2_{n-1}) \\ = (1+B_{n-1})^2 - (C_{n-1} + A_{n-1})^2 = 0 (the last equality holds if assuming 1 + A_{n-1} + B_{n-1} + C_{n-1} =0 ). Since 1 + A_1 + B_1 + C_1 =0 , from math induction, we obtain for any n, 1 + A_n + B_n + C_n = 0 .
It is clear now that fn(x) always has 1 as one of its roots. Let us denote D_n as the discriminant of the quadratic (x^2 + (A_n+1) x – C_n) . Then we should have (if fn(x) has three distinct roots)
- (claim 1) D_n >0 ;
- (claim 2) At x = 1, x^2 + (A_n+1) x - C_n = 2 + A_n - C_n \neq 0 .
Yet the two claims are not confirmed up to here. So more effort is needed! The following table (that lists feature related to function fn(x)) is made:
| An | Bn | Cn | ||||
| n=1 | -6 | 10 | -5 | 5>0 | 1 | |
| n=2 | -16 | 40 | -25 | 125>0 | 11 | |
| n=3 | -176 | 800 | -625 | 1752+4(-625)>0 | 451 |
By inspection, we find an interesting pattern that A_n <0, B_n >0, C_n <0 , or equivalently -A_n >0, B_n >0, -C_n >0 . From this we can deduce claim 2: 2 + A_n - C_n >2 >0 . We observe another interesting pattern: B_n > - C_n - 2 C_{n-1} >0 which implies B_n + C_n > -2 C_{n-1}>0, \\ -1 - A_n > -2 C_{n-1} >0. Note that (– Cn )= (Cn-1)2, and squaring on both sides of the previous line, to obtain: D_n = (1+A_n)^2 -4 (-C_n) >0 which is exactly Claim 1. (There is one last thing: it remains to be shown formally, the two patterns mentioned above .. ..)
From September 26th to November 14th, on most Saturdays (except a couple of Saturdays off), Access Academy welcome those students who desire to challenge math problems to the following sessions, that prepare them to participate in the Math Contests (problems given by University of Waterloo), and to write in the COMC (Canadian Open Math Challenge).
Each training session for COMC is 2.5 hours long. And each training session for CSMC (Canadian Senior Math Contests) or CIMC (Canadian Intermediate Math Contest) is 2 hours long.
While these training sessions are going, relevant posts will be added to the column, and we allow students to download some materials for training (these materials are password-protected). The posts are public accessible. More posts or training materials will be added when they are available.
See the relevant posts:
See the relevant training material:
公告板 – 数学竞赛报名消息
- 2026 年COMC 加拿大公开数学挑战赛征集报名(九月开始);CSMC/CIMC 加拿大高级与中级数学竞赛也征求报名。有意参加的学生可以联系 Access Academy。请注意相关的信息更新。
- 如果您对于加拿大数学竞赛不熟悉,请参考这个帖子。
- 2026 年的 CSMC/CIMC 将在11月18日(周三)举行。参加CSMC(高级数学竞赛)的应为在11或12 年级就读的学生。参加CIMC 的应是 9或10 年级就读的学生。
- 今年的COMC 加拿大公开数学挑战赛将在十月的最后一个周四(10月29日)举办。建议热爱数学挑战的学生参加。参赛学生应在 7 – 12 年级就读。COMC 是加拿大数学竞赛中接受公开报名的最高级别的挑战,也是唯一一个参赛者可能加入加拿大数学奥林匹克(CMO)及国际数学奥林匹克(IMO)的比赛(经选拔参加后续竞赛活动)。
- 今夏在上海举办的国际数学奥林匹克(IMO)上,中国,美国队获冠亚军。加拿大队取得第 13 名。
Prime numbers
Prime numbers are those that have 1 (one) and itself as the only two divisors. Examples of primes are 2, 3, 5, 7, 11. None of 4, 6, 9 is a prime since 4 = 2 × 2, 6 = 2 × 3, and 9 = 3 × 3.
If a number greater than one is not a prime, then it is a composite number, and can be factored into the product of primes — called prime factorization. We have given the prime factorization of 4, 6, 9 as above. For a couple of more examples:
12 = 2 × 2 × 3
36 = 2 × 3 × 3 × 3
28 = 2 × 2 × 7
So all natural numbers are divided into three classes: the number 1, the prime numbers, and the composite numbers.
A bonus point: π, besides representing in a circle, the ratio of circumference to diameter, also stands for a special function related to prime numbers. This is described as follows.
Function π(x) — for integer x, represents the number of primes less than or equal (i.e. not exceeding) x.
For example,
π(2) = 1, π(3) = 2, π(10) = 4, π(20) = 8 etc.
[To find why π(10) = 4, recall the 4 prime numbers not exceeding 10: they are 2,3,5, and 7.]