Skip to main content
NSWMaths Extension 1Syllabus dot point

When every outcome is equally likely, how do you turn a probability into a ratio of two counts, and how do the counting moves (multiplication principle, permutations, combinations, complement, cases) supply both the favourable count and the total?

Compute probabilities of equally likely outcomes as P = (favourable outcomes)/(total outcomes) where both counts come from the counting methods (multiplication principle, ^{n}P_{r} and ^{n}C_{r}), and apply the complement, separate-pool multiplication, cases and inclusion-exclusion to find the favourable count for selection, card-hand, digit-string and with/without-replacement problems

Turning counting into probability. When outcomes are equally likely, P = favourable/total and both counts come from the multiplication principle, nPr and nCr. Committee and card-hand probabilities by combinations, exactly-k of a type, at-least-one by the complement, a five-digit-number probability, inclusion-exclusion, and with versus without replacement.

Generated by Claude Opus 4.819 min answer

Reviewed by: AI editorial process; not yet individually human-reviewed

Have a quick question? Jump to the Q&A page

Jump to a section
  1. What this dot point is asking
  2. The answer
  3. How exam questions ask about this dot point

What this dot point is asking

The three counting pages of this module gave you machinery for counting outcomes: the multiplication principle and ordered selections for counting in stages with nrn^r and nPr^{n}P_{r}, the grouping, complement and cases moves for restricted arrangements, and combinations for unordered selections with nCr^{n}C_{r}. This page puts that machinery to work on probability. NESA's ME-A1 dot point asks you to compute a probability by counting: find the number of favourable outcomes and the number of total outcomes, then divide.

The single idea is the equally likely model. When every outcome of an experiment is equally likely, the probability of an event EE is

P(E)=number of outcomes in Enumber of outcomes in the sample space=n(E)n(S),P(E) = \frac{\text{number of outcomes in } E}{\text{number of outcomes in the sample space}} = \frac{n(E)}{n(S)},

and the whole skill is using counting to fill in those two numbers. Both the numerator and the denominator are counts of the same kind of object, so they must be counted the same way: if the sample space is "all 55-card hands", the favourable count must also be a count of hands. Everything else is choosing the right counting tool for each count, the multiplication principle, a permutation, a combination, the complement, cases, or inclusion-exclusion, and being consistent about ordered versus unordered between the two counts.

The answer

The equally likely model: probability as a ratio of two counts

A probability experiment has a set of possible outcomes called the sample space SS, and an event EE is a subset of those outcomes. When the outcomes are equally likely, no single outcome is favoured over another, so the probability of EE is simply the share of outcomes that lie in EE. Drawing a card "at random", dealing a hand "at random", choosing a committee "at random", arranging letters "in random order": each phrase signals that the underlying outcomes are equally likely, which is exactly the licence to count.

The reason the formula works is a direct sum. Each of the n(S)n(S) outcomes carries the same probability 1n(S)\dfrac{1}{n(S)}, and the event EE contains n(E)n(E) of them, so

P(E)=1n(S)+1n(S)++1n(S)n(E) terms=n(E)n(S).P(E) = \underbrace{\frac{1}{n(S)} + \frac{1}{n(S)} + \cdots + \frac{1}{n(S)}}_{n(E) \text{ terms}} = \frac{n(E)}{n(S)}.

So a probability question becomes two counting questions: count the sample space, and count the favourable event. The art is choosing what an "outcome" is and counting consistently.

Ordered or unordered? Choose once, and use it for both counts

Before counting anything, settle the two questions that run through all of combinatorics: are the outcomes ordered or unordered, and, if ordered, is repetition allowed? The answer decides which tool counts the sample space, and you must then count the favourable event the same way. Both choices are valid as long as they match, because any consistent factor cancels in the ratio.

A clean illustration is dealing 33 cards from a pack and asking for the probability of one club then two hearts in that order. Counting ordered outcomes, the sample space is 52P3=52×51×50=132600^{52}P_{3} = 52 \times 51 \times 50 = 132\,600, and the favourable outcomes are club, heart, heart, which is 13×13×12=202813 \times 13 \times 12 = 2028, giving 2028132600=13850\dfrac{2028}{132\,600} = \dfrac{13}{850}. If instead the question asked for one club and two hearts in any order, the natural count is unordered: the sample space is 52C3=22100^{52}C_{3} = 22\,100 and the favourable count is 13C113C2=13×78=1014^{13}C_{1}\,^{13}C_{2} = 13 \times 78 = 1014, giving 101422100=39850\dfrac{1014}{22\,100} = \dfrac{39}{850}, three times larger because the club can sit in any of three positions. The rule of thumb: if a problem can be done either way, use unordered selections, because the numbers are smaller and the combinations are easier to handle.

Selection probabilities with combinations

The commonest equally likely model is a random selection: a committee, a team, a hand, a handful of discs. Order does not matter, so the sample space is a combination, and the favourable count is a combination (or a product of combinations from separate pools). The whole pattern is

P(event)=combinations that make the eventall combinations of that size.P(\text{event}) = \frac{\text{combinations that make the event}}{\text{all combinations of that size}}.

Take a committee of 44 chosen at random from 77 women and 55 men, 1212 people in all. The sample space is every committee of 44, namely 12C4=495^{12}C_{4} = 495 equally likely committees. To find P(all 4 are women)P(\text{all } 4 \text{ are women}), count the all-women committees, 7C4=35^{7}C_{4} = 35, and divide:

P(all women)=7C412C4=35495=7990.0707.P(\text{all women}) = \frac{^{7}C_{4}}{^{12}C_{4}} = \frac{35}{495} = \frac{7}{99} \approx 0.0707.

The same total 495495 serves every event for this committee. The diagram below shows how the 495495 committees split by the number of men, and where the favourable count of 3535 sits.

Counts feeding the probability: choosing a committee of fourA committee of four is chosen from seven women and five men, a total of C(12,4) which is 495 equally likely committees. The table lists how many committees have zero, one, two, three or four men: the counts are 35, 175, 210, 70 and 5, adding to 495. The favourable count for all four being women is 35, so the probability is 35 over 495, which simplifies to 7 over 99. Horizontal bars show the relative sizes of the five cases.P = favourable counts ÷ total countcommittee of 4 from 7 women + 5 men: C(12,4) = 495 equally likely committeescase (by number of men)count0 men (all 4 women)351 man, 3 women1752 men, 2 women2103 men, 1 woman704 men (all 4 men)5total495P(all 4 women)35 / 495 = 7/99favourable = 35 (the top row);all 5 case counts add to 495.

The bars make the shape of the problem visible: most committees are mixed, the all-women and all-men committees are rare, and the five case counts 35+175+210+70+5=49535 + 175 + 210 + 70 + 5 = 495 recover the whole sample space. That last fact is a free check: the favourable counts of all the disjoint cases must add to the total.

Exactly kk of a type, and the separate-pools rule

When an event fixes how many come from each pool, choose from each pool separately and multiply the combinations, then divide by the total. Staying with the committee of 44 from 77 women and 55 men, the probability of exactly 22 men chooses 22 of the 55 men and 22 of the 77 women:

P(exactly 2 men)=5C27C212C4=10×21495=210495=14330.4242.P(\text{exactly } 2 \text{ men}) = \frac{^{5}C_{2}\,^{7}C_{2}}{^{12}C_{4}} = \frac{10 \times 21}{495} = \frac{210}{495} = \frac{14}{33} \approx 0.4242.

The numerator multiplies because the two pools are chosen independently; the denominator is the same 495495. Running through every value of "number of men" gives the full distribution, and the favourable counts 35,175,210,70,535, 175, 210, 70, 5 are exactly the column counts in the diagram above. Because these cases are disjoint and exhaustive, their probabilities sum to 11:

35+175+210+70+5495=495495=1,\frac{35 + 175 + 210 + 70 + 5}{495} = \frac{495}{495} = 1,

which is the most reliable check available on a "by cases" probability.

Card-hand probabilities

A standard pack has 5252 cards in 44 suits of 1313. A poker-style hand of 55 cards is an unordered selection, so the sample space is

52C5=2598960^{52}C_{5} = 2\,598\,960

equally likely hands, a number worth computing once and reusing. Every hand probability is then a combination count over 25989602\,598\,960. Three patterns cover most questions.

Exactly rr of one suit. For exactly 22 hearts, choose 22 of the 1313 hearts and the other 33 from the 3939 non-hearts:

P(exactly 2 hearts)=13C239C352C5=78×91392598960=7128422598960=9139333200.2743.P(\text{exactly } 2 \text{ hearts}) = \frac{^{13}C_{2}\,^{39}C_{3}}{^{52}C_{5}} = \frac{78 \times 9139}{2\,598\,960} = \frac{712\,842}{2\,598\,960} = \frac{9139}{33\,320} \approx 0.2743.

A specified split across two suits. For 33 spades and 22 hearts, choose from each suit and multiply:

P(3,2)=13C313C252C5=286×782598960=223082598960=143166600.0086.P(3\spadesuit, 2\heartsuit) = \frac{^{13}C_{3}\,^{13}C_{2}}{^{52}C_{5}} = \frac{286 \times 78}{2\,598\,960} = \frac{22\,308}{2\,598\,960} = \frac{143}{16\,660} \approx 0.0086.

All five of one suit (a flush). Any one of the 44 suits supplies all 55 cards, so multiply by 44:

P(all one suit)=4×13C552C5=4×12872598960=51482598960=33166600.0020.P(\text{all one suit}) = \frac{4 \times {}^{13}C_{5}}{^{52}C_{5}} = \frac{4 \times 1287}{2\,598\,960} = \frac{5148}{2\,598\,960} = \frac{33}{16\,660} \approx 0.0020.

Each is favourable-over-total with both counts as combinations of 55 cards, the only difference being how the favourable hands are built.

"At least one" by the complement

"At least one" almost always fragments into several cases ("exactly one", "exactly two", ...), so it is far shorter to count the complement, "none", and subtract its probability from 11:

P(at least one)=1P(none).P(\text{at least one}) = 1 - P(\text{none}).

Deal 55 cards and ask for P(at least one ace)P(\text{at least one ace}). The complement is a hand with no ace, drawn from the 4848 non-aces, 48C5=1712304^{48}C_{5} = 1\,712\,304, so

P(no ace)=48C552C5=17123042598960=35673541450.6588,P(\text{no ace}) = \frac{^{48}C_{5}}{^{52}C_{5}} = \frac{1\,712\,304}{2\,598\,960} = \frac{35\,673}{54\,145} \approx 0.6588,

P(at least one ace)=13567354145=18472541450.3412.P(\text{at least one ace}) = 1 - \frac{35\,673}{54\,145} = \frac{18\,472}{54\,145} \approx 0.3412.

Counting "at least one ace" directly would mean adding the exactly-one, exactly-two, exactly-three and exactly-four ace cases; the complement collapses all of that into one clean count.

A five-digit-number probability, and inclusion-exclusion

Digit problems use the multiplication principle for both counts. A five-digit number chosen at random has a first digit from 11 to 99 (99 choices, since a number cannot start with 00) and four further digits each from 00 to 99, so the sample space is

9×104=900009 \times 10^4 = 90\,000

equally likely numbers. For P(at least one 4)P(\text{at least one } 4), reach for the complement: a number with no 44 has 88 choices for the first digit (11 to 99 except 44) and 99 for each of the other four (00 to 99 except 44), so 8×94=524888 \times 9^4 = 52\,488 have no 44, and the rest have at least one:

P(at least one 4)=900005248890000=3751290000=52112500.4168.P(\text{at least one } 4) = \frac{90\,000 - 52\,488}{90\,000} = \frac{37\,512}{90\,000} = \frac{521}{1250} \approx 0.4168.

The case tree shows the split: the whole sample space branches into "no 44" and "at least one 44", and the favourable count is what is left after removing the complement.

Complement split for at least one 4 in a five-digit numberA five-digit number is chosen at random. The first digit is one of nine choices (1 to 9) and each of the other four digits is one of ten, so there are 9 times 10 to the fourth, which is 90000 numbers in all. This splits into two cases: numbers with no 4 at all, counted by 8 times 9 to the fourth which is 52488, and numbers with at least one 4, which is the rest, 90000 minus 52488 equals 37512. The probability of at least one 4 is 37512 over 90000, which simplifies to 521 over 1250.Counting "at least one 4" by its complementall 5-digit numbers900009 × 10⁴no 4 at all (complement)524888 × 9⁴at least one 4 (favourable)3751290000 − 52488P(at least one 4) = 37512 / 90000 = 521/1250

When two "at least one" conditions are joined by "and", the complement of the event is an "or", and an "or" of overlapping sets needs inclusion-exclusion. For P(at least one 4 and at least one 5)P(\text{at least one } 4 \textbf{ and } \text{at least one } 5), let AA be the numbers with no 44 and BB the numbers with no 55. Then the numbers failing the event (no 44 or no 55) form ABA \cup B, and

n(AB)=n(A)+n(B)n(AB).n(A \cup B) = n(A) + n(B) - n(A \cap B).

Here n(A)=n(B)=8×94=52488n(A) = n(B) = 8 \times 9^4 = 52\,488, and ABA \cap B (no 44 and no 55) has 77 choices for the first digit and 88 for each of the rest, 7×84=286727 \times 8^4 = 28\,672. So

n(AB)=52488+5248828672=76304,n(A \cup B) = 52\,488 + 52\,488 - 28\,672 = 76\,304,

and the favourable numbers (at least one 44 and at least one 55) are everything else:

P(1 four and1 five)=900007630490000=1369690000=85656250.1522.P(\ge 1 \text{ four and} \ge 1 \text{ five}) = \frac{90\,000 - 76\,304}{90\,000} = \frac{13\,696}{90\,000} = \frac{856}{5625} \approx 0.1522.

The key move is recognising that "and" on two at-least-one conditions complements to an "or", and that the "or" overlaps, so the overlap ABA \cap B must be subtracted once.

With replacement versus without replacement

Whether an item is replaced between draws changes the model, and so changes both counts. Without replacement (or drawing all at once), the items drawn are distinct, so a selection of rr is naturally an unordered combination, and the probability is a ratio of combinations. With replacement, each draw is from the full set and the draws are independent, so probabilities multiply.

Take a bag of 33 red, 77 yellow and 55 blue balls (1515 in all) and draw 33. Without replacement, the sample space is 15C3=455^{15}C_{3} = 455 and P(all yellow)P(\text{all yellow}) uses the favourable count 7C3=35^{7}C_{3} = 35:

P(all yellow, no replacement)=7C315C3=35455=1130.0769.P(\text{all yellow, no replacement}) = \frac{^{7}C_{3}}{^{15}C_{3}} = \frac{35}{455} = \frac{1}{13} \approx 0.0769.

With replacement, each draw is yellow with probability 715\dfrac{7}{15} independently, so

P(all yellow, with replacement)=(715)3=34333750.1016.P(\text{all yellow, with replacement}) = \left(\frac{7}{15}\right)^3 = \frac{343}{3375} \approx 0.1016.

Replacement is slightly more likely here, because the proportion of yellow is held at 715\dfrac{7}{15} on every draw, whereas without replacement each yellow removed leaves a smaller proportion for the next draw. Spotting "drawn together" or "all at once" (no replacement, combinations) versus "replaced before the next" (with replacement, independent multiplication) is half the battle on these questions.

How exam questions ask about this dot point

The wording tells you the model and the move. Map the phrase to the method:

  • "...chosen / selected / drawn at random", "a committee / team / hand is formed". Equally likely outcomes: P=favourabletotalP = \dfrac{\text{favourable}}{\text{total}}, almost always with combinations.
  • "...all of one type", "all male", "all the same suit". Favourable is a single combination from one pool over the total: aCrnCr\dfrac{^{a}C_{r}}{^{n}C_{r}}.
  • "...exactly kk of one type and the rest another", "two men and three women". Multiply per-pool combinations in the numerator: aCibCjnCr\dfrac{^{a}C_{i}\,^{b}C_{j}}{^{n}C_{r}}.
  • "...at least one", "at most", "none ... so find the chance of some". Complement: 1P(none)1 - P(\text{none}).
  • "...at least one XX and at least one YY", "contains both". Complement to an "or", then inclusion-exclusion on the overlap.
  • "...a majority", "more Labor than Liberal", "at least 33". Split into the qualifying cases, count each by multiplying pools, then add.
  • "...a five-digit number / a four-digit code is chosen at random; find the probability that ...". Multiplication principle for both counts; mind the no-leading-zero restriction.
  • "...drawn together / all at once / without replacement" versus "...replaced before the next draw". Combinations for without replacement; independent multiplication for with replacement.

Practice questions

Original practice questions graded from foundation to exam level, each with a full worked solution. Try them before revealing the solution.

foundation3 marksA club has 99 members, of whom 55 are men and 44 are women. A committee of 33 is chosen at random. (i) How many different committees are possible? (ii) Find the probability that the committee consists entirely of men.
Show worked solution →

Part (i): the total is an unordered selection. A committee is unordered, so count combinations. Choose any 33 of the 99 members:

9C3=9×8×73!=5046=84.^{9}C_{3} = \frac{9 \times 8 \times 7}{3!} = \frac{504}{6} = 84.

Part (ii): the favourable count is the all-men committees. All 33 members come from the 55 men:

5C3=5×4×33!=10.^{5}C_{3} = \frac{5 \times 4 \times 3}{3!} = 10.

Form the ratio favourable over total. Both counts use the same equally likely committees, so

P(all men)=5C39C3=1084=5420.119.P(\text{all men}) = \frac{^{5}C_{3}}{^{9}C_{3}} = \frac{10}{84} = \frac{5}{42} \approx 0.119.

Answer. (i) 8484 committees; (ii) P(all men)=5420.119P(\text{all men}) = \dfrac{5}{42} \approx 0.119.

foundation3 marksThe integers 11 to 1010 are written on ten cards. Four cards are drawn at random (all at once). Find the probability that (i) the card numbered 99 is among those drawn, and (ii) the card numbered 88 is not drawn.
Show worked solution →

Count the total selections. Four cards chosen from ten, order irrelevant:

10C4=10×9×8×74!=504024=210.^{10}C_{4} = \frac{10 \times 9 \times 8 \times 7}{4!} = \frac{5040}{24} = 210.

Part (i): force card 99 in, then choose the rest. With 99 already drawn, the other 33 cards come from the remaining 99, giving 9C3=84^{9}C_{3} = 84 favourable selections:

P(9 drawn)=9C310C4=84210=25=0.4.P(9 \text{ drawn}) = \frac{^{9}C_{3}}{^{10}C_{4}} = \frac{84}{210} = \frac{2}{5} = 0.4.

Part (ii): ban card 88, then choose freely. Removing 88 leaves 99 cards, from which all 44 are drawn, giving 9C4=126^{9}C_{4} = 126 favourable selections:

P(8 not drawn)=9C410C4=126210=35=0.6.P(8 \text{ not drawn}) = \frac{^{9}C_{4}}{^{10}C_{4}} = \frac{126}{210} = \frac{3}{5} = 0.6.

Answer. (i) 25=0.4\dfrac{2}{5} = 0.4; (ii) 35=0.6\dfrac{3}{5} = 0.6.

core4 marksFive cards are dealt from a standard pack of 5252. Find the probability that (i) all five cards are hearts, and (ii) exactly one of the five cards is a heart. Leave each answer as a fraction and as a decimal to four places.
Show worked solution →

Count the total number of hands. A hand is an unordered selection of 55 from 5252:

52C5=2598960.^{52}C_{5} = 2\,598\,960.

Part (i): all five from the 1313 hearts. The favourable count is 13C5=1287^{13}C_{5} = 1287, so

P(all hearts)=13C552C5=12872598960=33666400.0005.P(\text{all hearts}) = \frac{^{13}C_{5}}{^{52}C_{5}} = \frac{1287}{2\,598\,960} = \frac{33}{66\,640} \approx 0.0005.

Part (ii): one heart and four non-hearts, multiplied. Choose 11 of the 1313 hearts and 44 of the 3939 non-hearts, then multiply (separate pools):

13C1×39C4=13×82251=1069263.^{13}C_{1} \times {}^{39}C_{4} = 13 \times 82\,251 = 1\,069\,263.

Hence

P(exactly one heart)=10692632598960=27417666400.4114.P(\text{exactly one heart}) = \frac{1\,069\,263}{2\,598\,960} = \frac{27\,417}{66\,640} \approx 0.4114.

Answer. (i) 33666400.0005\dfrac{33}{66\,640} \approx 0.0005; (ii) 27417666400.4114\dfrac{27\,417}{66\,640} \approx 0.4114.

core3 marksA five-digit number is chosen at random (so the first digit is not 00). Find the probability that all five of its digits are different. Leave the answer as a fraction and as a decimal to four places.
Show worked solution →

Count the total of five-digit numbers (the sample space). The first digit has 99 choices (11 to 99), and each of the other four has 1010 choices:

9×104=90000.9 \times 10^4 = 90\,000.

Count the favourable numbers slot by slot, without repetition. The first digit avoids 00, so 99 choices; the second avoids only the first used digit, so 99 choices (the 00 is now back in play); then 88, then 77, then 66:

9×9×8×7×6=27216.9 \times 9 \times 8 \times 7 \times 6 = 27\,216.

Form the ratio.

P(all digits different)=2721690000=189625=0.3024.P(\text{all digits different}) = \frac{27\,216}{90\,000} = \frac{189}{625} = 0.3024.

Why the second slot still has 99. The first digit removed one value but freed nothing; the digit 00, banned only from the first slot, becomes available from the second slot onward, so the falling pattern is 9,9,8,7,69, 9, 8, 7, 6 rather than 9,8,7,6,59, 8, 7, 6, 5.

Answer. 189625=0.3024\dfrac{189}{625} = 0.3024.

exam4 marksA committee of 55 is chosen at random from 66 Labor and 55 Liberal members of a council. Find the probability that Labor holds a majority on the committee (that is, at least 33 of the 55 members are from Labor). Leave the answer as a fraction and as a decimal to four places.
Show worked solution →

Count the total committees. Five chosen from 1111, order irrelevant:

11C5=462.^{11}C_{5} = 462.

Read 'majority' as the cases 33, 44 or 55 Labor, and add them. Each case fixes how many come from each pool, so multiply within a case and add across cases. The favourable count is

6C35C2+6C45C1+6C55C0.^{6}C_{3}\,^{5}C_{2} + {}^{6}C_{4}\,^{5}C_{1} + {}^{6}C_{5}\,^{5}C_{0}.

Evaluate each case.

6C35C2=20×10=200,6C45C1=15×5=75,6C55C0=6×1=6.^{6}C_{3}\,^{5}C_{2} = 20 \times 10 = 200, \quad ^{6}C_{4}\,^{5}C_{1} = 15 \times 5 = 75, \quad ^{6}C_{5}\,^{5}C_{0} = 6 \times 1 = 6.

Add and form the ratio. The favourable total is 200+75+6=281200 + 75 + 6 = 281, so

P(Labor majority)=2814620.6082.P(\text{Labor majority}) = \frac{281}{462} \approx 0.6082.

Answer. 2814620.6082\dfrac{281}{462} \approx 0.6082.

exam4 marksA bag holds 77 white and 55 black discs. Three discs are taken. Find the probability that all three are black if (i) the discs are taken without replacement (all at once), and (ii) the discs are taken one at a time, each replaced before the next is drawn. Leave each answer as a fraction and as a decimal to four places.
Show worked solution →

Part (i): without replacement is an unordered selection. The total is 12C3=220^{12}C_{3} = 220 and the favourable count is the all-black selections 5C3=10^{5}C_{3} = 10:

P(3 black, no replacement)=5C312C3=10220=1220.0455.P(\text{3 black, no replacement}) = \frac{^{5}C_{3}}{^{12}C_{3}} = \frac{10}{220} = \frac{1}{22} \approx 0.0455.

Part (ii): with replacement, each draw is independent. The bag is restored each time, so each draw is black with probability 512\dfrac{5}{12}, and the three draws multiply:

P(3 black, with replacement)=(512)3=12517280.0723.P(\text{3 black, with replacement}) = \left(\frac{5}{12}\right)^3 = \frac{125}{1728} \approx 0.0723.

Compare the two. Replacement keeps 55 black among 1212 every time, so it is slightly more likely to draw three blacks than the without-replacement case, where each black removed leaves fewer for the next draw.

Answer. (i) 1220.0455\dfrac{1}{22} \approx 0.0455; (ii) 12517280.0723\dfrac{125}{1728} \approx 0.0723.

Related dot points