This page is the workout room for combinations . The parent note built the machinery; here we drag it through every kind of situation a problem can hand you — tiny edge cases, degenerate zeros, symmetry shortcuts, "at least one" complements, real-world word problems, and an exam-style trap.
Before we count anything, one reminder in plain words: ( r n ) (say "n choose r ") is how many different groups of r things you can pull out of n things when the order you grabbed them in does not matter . Every symbol below is that same idea in disguise.
Think of a topic as a machine with dials. Each dial setting is a different kind of problem. If we test every setting, nothing on the exam can surprise us. Here are the settings for combinations:
#
Case class
What makes it tricky
Hit by example
A
Ordinary 0 < r < n
plain counting, expect a medium number
Ex 1
B
Edge: r = 0 (choose nothing)
tempting to say 0 , truth is 1
Ex 2
C
Edge: r = n (choose everything)
only one full group exists
Ex 2
D
Degenerate: r > n (asking too much)
impossible — answer must be 0
Ex 2
E
Symmetry shortcut r close to n
huge-looking, actually tiny via ( r n ) = ( n − r n )
Ex 3
F
Pascal check — build a value two ways
verifies the addition law
Ex 4
G
"At least one" via complement
many sub-cases collapse to Total − None
Ex 5
H
Split selection (choose from two groups)
multiply independent choices
Ex 6
I
Real-world word problem
translate English → n and r
Ex 7
J
Exam twist : permutation-vs-combination trap
order secretly matters (or doesn't)
Ex 8
We now walk through examples 1–8 so that every row above gets covered .
Worked example Choose a 4-player team from 9 players
How many different 4-player teams can be formed from 9 distinct players?
Forecast: guess before reading — will it be more or fewer than 100 ? Jot a number.
Step 1. Recognise: a team has no internal order — swapping two chosen players gives the same team.
Why this step? Combinations, not permutations, exactly because order is irrelevant. If we cared who is "captain first," it would be a permutation .
Step 2. Write the falling-factor form (top r = 4 factors of 9 , over 4 ! ):
( 4 9 ) = 4 ! 9 ⋅ 8 ⋅ 7 ⋅ 6 = 24 9 ⋅ 8 ⋅ 7 ⋅ 6 .
Why this step? The numerator 9 ⋅ 8 ⋅ 7 ⋅ 6 is the ordered count 9 P 4 ; dividing by 4 ! = 24 erases the ordering we don't want (the "un-shuffle").
Step 3. Simplify: 24 3024 = 126.
Verify: Cross-check with the symmetry ( 4 9 ) = ( 5 9 ) ; and against Pascal's row n = 9 : 1 , 9 , 36 , 84 , 126 , 126 , 84 , 36 , 9 , 1 — the fifth entry is indeed 126 . ✓
( 0 6 ) , ( 6 6 ) , and ( 8 6 )
Evaluate all three.
Forecast: which of these do you think equals 0 ?
Step 1 (Cell B, r = 0 ). "How many ways to choose nothing from 6?" There is exactly one way: take the empty set.
( 0 6 ) = 0 ! 6 ! 6 ! = 1 ⋅ 6 ! 6 ! = 1.
Why this step? The value hinges on $0!=1$ . If you wrongly thought 0 ! = 0 you'd divide by zero — nonsense. There's genuinely one empty group, so 1 is right.
Step 2 (Cell C, r = n ). "Choose all 6 from 6." Only one full group exists.
( 6 6 ) = 6 ! 0 ! 6 ! = 1.
Why this step? Same 0 ! = 1 engine — and it matches Step 1 by symmetry ( 0 6 ) = ( 6 6 ) .
Step 3 (Cell D, r > n ). "Choose 8 from 6" is impossible — you can't pull more items than exist.
( 8 6 ) = 0.
Why this step? In the formula 8 ! ( 6 − 8 )! 6 ! the term ( 6 − 8 )! = ( − 2 )! is undefined, which is nature's way of screaming "impossible." The count of impossible things is 0 .
Verify: The whole n = 6 Pascal row is 1 , 6 , 15 , 20 , 15 , 6 , 1 — its ends are 1 (matching Steps 1–2), and there is no 9th entry (matching Step 3's 0 ). ✓
( 48 50 )
Forecast: does computing 50 ! sound fun? It shouldn't.
Step 1. Use the symmetry identity ( r n ) = ( n − r n ) with n = 50 , r = 48 :
( 48 50 ) = ( 2 50 ) .
Why this step? Choosing 48 people to keep is the same decision as choosing the 2 to leave out — one choice, two names. Always convert to the smaller r to minimise multiplications.
Step 2. Now it's trivial:
( 2 50 ) = 2 50 ⋅ 49 = 2 2450 = 1225.
Why this step? Only two falling factors on top (50 ⋅ 49 ), divide by 2 ! = 2 .
Verify: Both sides must be equal by the identity, and 1225 is a triangular number 2 50 ⋅ 49 — indeed ( 2 n ) is always the ( n − 1 ) th triangular number. ✓
( 3 6 ) from its two Pascal parents — with a picture
Forecast: which two numbers in row n = 5 do you think add to give it?
Step 1. Compute directly first: ( 3 6 ) = 3 ! 6 ⋅ 5 ⋅ 4 = 6 120 = 20.
Why this step? We want a target to confirm the addition law reproduces.
Step 2. Apply Pascal's rule ( r n ) = ( r − 1 n − 1 ) + ( r n − 1 ) with n = 6 , r = 3 :
( 3 6 ) = ( 2 5 ) + ( 3 5 ) .
Why this step? Story-proof (from the parent): fix one special person "Ravi." Either Ravi is IN the committee — then pick r − 1 = 2 more from the other 5 , giving ( 2 5 ) — or Ravi is OUT — pick all 3 from the other 5 , giving ( 3 5 ) . These two cases don't overlap and cover everything, so we add . The red cell in the figure is the sum of the two black cells leaning over it.
Step 3. Evaluate the parents: ( 2 5 ) = 10 and ( 3 5 ) = 10 , so 10 + 10 = 20.
Verify: 20 = 20 ✓ — the direct value equals the Pascal sum. This is exactly the addition law that grows the whole triangle.
Worked example From 5 physicists and 3 chemists, choose 4 scientists with
at least one chemist
Forecast: would you rather add up "1 chemist" + "2 chemists" + "3 chemists" separately, or do one subtraction?
Step 1. Count all groups of 4 from the 5 + 3 = 8 scientists:
( 4 8 ) = 4 ! 8 ⋅ 7 ⋅ 6 ⋅ 5 = 24 1680 = 70.
Why this step? "At least one chemist" is messy (many cases), but its opposite — "no chemist at all" — is a single clean count. Start with the whole universe.
Step 2. Count the complement : groups with zero chemists, i.e. all 4 chosen from the 5 physicists:
( 4 5 ) = ( 1 5 ) = 5.
Why this step? Used symmetry ( 4 5 ) = ( 1 5 ) to make it instant. This is the "none" case we will remove.
Step 3. Subtract:
70 − 5 = 65.
Why this step? Total groups minus the ones we don't want (no chemist) leaves exactly the "at least one chemist" groups.
Verify (the long way, must agree): count by exact number of chemists:
1 chemist: ( 1 3 ) ( 3 5 ) = 3 ⋅ 10 = 30
2 chemists: ( 2 3 ) ( 2 5 ) = 3 ⋅ 10 = 30
3 chemists: ( 3 3 ) ( 1 5 ) = 1 ⋅ 5 = 5
Sum = 30 + 30 + 5 = 65. ✓ Matches the complement method.
Worked example A committee of 2 boys AND 3 girls from 5 boys and 6 girls
Forecast: do we add the two counts or multiply them?
Step 1. Choose the boys, ignoring girls: ( 2 5 ) = 2 5 ⋅ 4 = 10.
Why this step? The boy-choice is a self-contained combination; order among boys is irrelevant.
Step 2. Choose the girls, independently: ( 3 6 ) = 6 6 ⋅ 5 ⋅ 4 = 20.
Why this step? Same logic for girls, and this choice does not interact with which boys were picked.
Step 3. Combine by the multiplication principle :
( 2 5 ) ⋅ ( 3 6 ) = 10 ⋅ 20 = 200.
Why this step? For each of the 10 boy-groups, all 20 girl-groups are possible, so total pairs = 10 × 20 . We multiply (independent stages), unlike Example 5 where disjoint cases were added .
Verify: Units sanity — we wanted committees, and 10 (boy-ways) × 20 (girl-ways) = 200 full committees. Every committee has exactly 2 boys and 3 girls as required, none double-counted. ✓
Worked example Pizza toppings
A shop offers 8 distinct toppings . A "supreme" pizza uses exactly 3 of them, and toppings can't repeat. How many different supreme pizzas exist? Does adding pepperoni first vs last matter?
Forecast: guess whether it's under or over 60 .
Step 1. Translate English → numbers: "8 distinct toppings" ⇒ n = 8 ; "exactly 3, no repeats" ⇒ r = 3 ; "a pizza doesn't remember the order you sprinkled toppings" ⇒ order irrelevant ⇒ combination.
Why this step? The phrase "order doesn't matter" is the whole decision between ( 3 8 ) and 8 P 3 . A pizza with {ham, olive, chili} is identical to {chili, olive, ham}.
Step 2. Compute:
( 3 8 ) = 3 ! 8 ⋅ 7 ⋅ 6 = 6 336 = 56.
Why this step? Top three falling factors over 3 ! — the standard un-shuffle.
Verify: Symmetry cross-check ( 3 8 ) = ( 5 8 ) ; and Pascal's rule ( 3 8 ) = ( 2 7 ) + ( 3 7 ) = 21 + 35 = 56. ✓ Two independent checks agree.
Worked example Same numbers, different question — spot the switch
From 8 runners, (a) how many ways to pick a group of 3 to advance to the final, and (b) how many ways to award gold, silver, bronze medals to 3 of them?
Forecast: will (a) and (b) be equal? If not, which is bigger and by what factor?
Step 1 (a) — group, no order. Advancing is a plain committee:
( 3 8 ) = 3 ! 8 ⋅ 7 ⋅ 6 = 56.
Why this step? Whether runner X advances "1st-named" or "3rd-named" makes no difference — same trio advances. Order irrelevant ⇒ combination.
Step 2 (b) — medals, order matters. Gold ≠ silver ≠ bronze, so the arrangement of the same three runners now gives different outcomes. This is a permutation :
8 P 3 = 8 ⋅ 7 ⋅ 6 = 336.
Why this step? Here swapping who gets gold and who gets bronze is a genuinely different result — so we must not divide by 3 ! . This is exactly the trap in the parent note's "Error A," run in reverse.
Step 3 — see the relationship. The two answers differ by precisely the un-shuffle factor:
8 P 3 = ( 3 8 ) ⋅ 3 ! = 56 ⋅ 6 = 336.
Why this step? Every unordered trio (56 of them) can be medal-ordered in 3 ! = 6 ways, giving 336 . This is the master relation ( r n ) = r ! n P r , read forwards.
Verify: 56 ⋅ 6 = 336 ✓, and 336/6 = 56 ✓. The single test — "does swapping two chosen people change the answer?" — cleanly separates (a) from (b).
Common mistake The one question that resolves every case
Before computing anything, ask: "Does swapping two chosen items give a different result?"
No ⇒ combination ( r n ) (committee, hand, team, toppings, subset).
Yes ⇒ permutation n P r (medals, rankings, seating, passwords).
And for "at least one," flip to the complement; for split groups, multiply ; for near-full r , use symmetry .
Why is ( 0 6 ) = 1 and not 0 ? There is exactly one way to choose nothing (the empty set); the formula gives 0 ! 6 ! 6 ! = 1 using 0 ! = 1 .
What is ( 8 6 ) and why? 0 — you cannot choose more items than exist; ( 6 − 8 )! is undefined, signalling impossibility.
Fast way to compute ( 48 50 ) ? Use symmetry ( 48 50 ) = ( 2 50 ) = 2 50 ⋅ 49 = 1225 .
"At least one chemist" from a pool — strategy? Total minus complement: ( 4 8 ) − ( 4 5 ) = 70 − 5 = 65 .
2 boys and 3 girls — add or multiply the sub-counts? Multiply (independent stages): ( 2 5 ) ( 3 6 ) = 10 ⋅ 20 = 200 .
Group of 3 vs gold/silver/bronze from 8 — which is bigger and by how much? Medals bigger by factor 3 ! : 8 P 3 = 336 = 6 ⋅ ( 3 8 ) = 6 ⋅ 56 .
Mnemonic The matrix in one breath
Zero-choose = 1, over-choose = 0, symmetry shrinks r , "at least one" subtracts the "none," split groups multiply, medals don't un-shuffle.
Combinations — nCr, Pascal's triangle — the parent machinery these examples exercise.
Permutations — nPr — Example 8's trap lives on the order-matters side.
Factorials and 0! — Example 2's edge cases hinge on 0 ! = 1 .
Binomial Theorem — the same ( r n ) appear as coefficients of ( a + b ) n .
Probability — counting outcomes — these counts become numerators/denominators of probabilities.
Pascal's Triangle patterns — Example 4's addition law is the triangle's growth rule.