JC1 H3 Mathematics Tutor Bukit Timah | Circular Permutations, Rotation and Reflection. Did you know? Six people can stand in 720 different lines, yet there are only 120 different seatings at a round table when turning the table does not create a new arrangement. The difference comes from deciding which transformations count as the same outcome.
This guide is designed for JC1 Mathematics students near Bukit Timah, Sixth Avenue and Beauty World strengthening H2 probability and combinatorial reasoning for potential later H3 study. We distinguish ordinary permutations, rotations, reflections, repeated colours and conditional seating rules before using counting arguments and exact examples.
The 2027 H3 Mathematics syllabus 9820 assumes H2 probability and counting, and explicitly includes the symmetry principle, combinatorial arguments and the bijection principle. Circular permutations are a natural instructional application of those skills, not a separately promised examination topic. These are original teaching problems; the student’s junior college determines H3 subject availability.
The quick answer: rotate, reflect or label?
- In a line: n distinct labelled objects have n! arrangements.
- Around an unlabelled round table: if rotations are identified and all n people are distinct, there are (n−1)! arrangements.
- Reflections: ordinary round-table seating generally treats mirror orders as different. If the question explicitly identifies mirror images, divide by two again for n≥3 distinct objects.
- Numbered seats: if the positions are physically labelled, rotating everyone generally changes who occupies each position, so the total is n! unless further rules say otherwise.
- Repeated colours: blindly dividing a linear arrangement count by n may fail because some circular patterns have rotational symmetry. Use an orbit argument or Burnside’s counting method.
The first job is to define what constitutes one outcome. Are we counting the people’s relative order, which seats each person occupies, or patterns of indistinguishable beads? A correct counting formula can be wrong for the intended experiment if those objects are confused.
Why (n−1)! works for distinct people
Arrange n distinct people in a line. There are n! orders. Every circular seating has n descriptions obtained by starting the reading at each of its n people, and these n descriptions are all distinct because the people are distinct. Therefore dividing n! by n gives (n−1)!.
Another proof fixes one named person, say Alex, at an arbitrary reference position. The other n−1 distinct people can then be placed in (n−1)! orders clockwise. These two explanations express the same rotational symmetry in different ways.
Worked examples: circular arrangements with distinct people
Worked example 1: Four people around one round table
With four distinct people A,B,C,D, the number of circular seating orders up to rotation is (4−1)!=3!=6. Fix A at a reference position and permute B,C,D in the other three places.
The linear count 4!=24 would count each circular seating four times, corresponding to choosing four different starting points around the same circle. Reflection is not being identified, so clockwise and anticlockwise orders may differ.
Worked example 2: Six distinct people around a table
For six distinct people, the number of circular arrangements up to rotation is (6−1)!=5!=120. Fixing one person at the top leaves five independently ordered positions.
It is not six factorial because the table has no preferred starting chair in the model. If the six seats were explicitly numbered, however, the total would instead be 6!=720. The physical description matters.
Worked example 3: Seven people and a fixed reference person
A round table has seven distinct diners and no numbered seats. Fix one chosen diner at the north reference point. Arrange the remaining six around them in 6!=720 ways.
This fixing method is legitimate because rotating the entire arrangement creates no new relative order. It is not the same as treating one diner as absent; that person remains part of the circular seating.
Worked example 4: Two named diners must sit together
Among six distinct people, suppose Alex and Bea must be adjacent. Treat them as one ordered block, which can appear as AB or BA. There are now five circular objects.
The number is (5−1)!×2=4!×2=48. The block method works because the pair occupies neighbouring seats and each valid arrangement corresponds to exactly one oriented pair-block seating.
Worked example 5: Two named diners may not sit together
Among six distinct people, the total circular arrangements are 120. Exactly 48 place Alex and Bea adjacent, by the block argument.
Therefore the count with them nonadjacent is 120−48=72. This complement method is valid because every seating falls into precisely one of the two categories: adjacent or nonadjacent.
Worked example 6: Opposite seats at an even table
Six distinct people sit around a circle. Fix Alex at a reference seat. There is exactly one seat opposite Alex, and Bea is required to occupy it.
The remaining four people may occupy the other four positions in 4!=24 ways. The count depends on there being an even number of equally spaced seats and one uniquely opposite position.
Worked example 7: Probability of adjacency
If every circular seating of six distinct people is equally likely, the probability that two named diners are neighbours is 48/120=2/5.
Equivalently, fix one diner and note that the second has two neighbouring seats among five remaining seats, giving 2/5. The direct probability argument confirms the combinatorial block calculation.
Worked example 8: Probability of an opposite seat
For six distinct people in equally likely circular arrangements, fix one person. Of the five remaining positions, exactly one is opposite.
Therefore the chance that a particular second person sits opposite is 1/5, matching the count 24 favourable arrangements out of 120. This condition is defined only when the seating geometry has a unique opposite position.
Worked example 9: All three couples must sit together
Six distinct people form three named couples, and each couple must occupy adjacent seats. Treat each couple as one block around the table.
There are (3−1)!=2 circular orders of the three blocks and 2³=8 choices of within-couple order, giving 2×8=16 arrangements. Reflection is still considered different in this standard round-table count.
Worked example 10: Four labelled couples all adjacent
Eight distinct people form four named couples, each required to sit together. The four blocks have (4−1)!=6 circular orders and each block has two internal orientations.
The total is 6×2⁴=96 arrangements. Without adjacency restrictions there would be 7!=5040 circular arrangements, so the constraint is substantial.
Worked example 11: Alternating men and women
Suppose three distinct men and three distinct women sit around a circle, alternating gender. Arrange the three women in (3−1)!=2 circular orders, then place the men in the three gaps in 3!=6 orders.
This yields 12 alternating arrangements. Choosing men first gives the same result, not another factor of two; both descriptions count the same alternating seatings.
Worked example 12: A fixed clockwise order
Suppose A,B,C are three distinct diners and the requirement says they must appear in clockwise order, though not necessarily consecutively, among five people.
Fix A as a reference. Of the 4!=24 arrangements of the other people, exactly half place B before C when travelling clockwise from A. Thus there are 12. Swapping B and C pairs the allowed and disallowed outcomes.
Worked example 13: Clockwise and anticlockwise considered different
For four distinct people A,B,C,D, the circular arrangement A-B-C-D in clockwise order is different from A-D-C-B if only rotations are ignored.
The two lists are mirror images but not rotationally equivalent when all labels are distinct. Therefore the standard circular count is six rather than three. A question must explicitly identify reflections before dividing by two.
Worked example 14: Mirrored necklaces of distinct beads
If a bracelet with four distinct beads is considered identical after rotation or reflection, the six rotational arrangements form three mirror pairs.
There are therefore 3 different bracelets. For n≥3 distinct labelled beads, no arrangement is fixed by a reflection, so identifying mirror images divides the usual (n−1)! circular count by two.
Worked example 15: When seats are labelled
A circular venue has four positions numbered 1,2,3,4. Four distinct guests assigned to these numbered seats have 4!=24 assignments.
Rotating all guests moves each person to a different numbered seat, so the outcome changes. One should not use (n−1)! unless rotations are truly equivalent under the problem’s definition.
Worked example 16: A round table with one marked chair
Even if the table is round, a single visibly marked chair provides a fixed reference position. For four distinct diners with no further restrictions, there are 4!=24 assignments to the marked and unmarked relative seats.
This is not the same as an unmarked circular seating, where rotations are equivalent. The physical mark removes the rotational symmetry that justified dividing by four.
Worked example 17: Five people, one adjacent pair forbidden
For five distinct diners, the total circular arrangements are 4!=24. If A and B are adjacent, combining them as one block gives 2×(4−1)!=12 arrangements.
So the number with A and B not adjacent is 24−12=12. The probability of adjacency here is 1/2 because each person has two neighbouring positions among four available positions relative to a fixed diner.
Worked example 18: Three people and rotational equivalence
With distinct A,B,C around an unmarked circle, there are (3−1)!=2 clockwise orders. These are A-B-C and A-C-B when A is fixed.
They are reflections of one another, not rotations. If mirror reflection is also considered equivalent, these two orders collapse into one bracelet arrangement.
Worked example 19: Anchor one person carefully
Fixing person A at the top is a valid symmetry-breaking device only when the table has no meaningful absolute orientation. If there is a stage facing north and seat location matters, moving everyone may change the situation.
Always state the model before anchoring. Otherwise a student may divide by n in an experiment where each labelled seat is a distinct possible outcome.
Worked example 20: A general adjacent-pair formula
For n≥3 distinct people around an unmarked circle, the number with two specified people adjacent is 2(n−2)!. Treat the pair as a block among n−1 circular objects.
The probability under equally likely circular arrangements is 2(n−2)!/(n−1)!=2/(n−1). This general result agrees with 2/5 for n=6 and 1/2 for n=5.
Worked example 21: A complement formula for nonadjacency
With n≥3 distinct people, the total is (n−1)! and adjacency count is 2(n−2)!. The count for two specified people not adjacent is (n−1)!−2(n−2)!.
For n=4 this is 6−4=2, corresponding to the two orders in which the pair sit opposite. The formula uses the standard unmarked-circle, distinct-person model.
Worked example 22: A probability with no collision
Choose a circular seating of n≥3 distinct people uniformly among its (n−1)! orders. The probability two given people are not adjacent is 1−2/(n−1)=(n−3)/(n−1).
The result is zero when n=3 because every pair on a three-seat circle are neighbours. Checking the smallest case catches indexing errors in the general formula.
Worked example 23: A labelled set of five seat positions
If five seats lie in a circle and are numbered, assigning five distinct people gives 5!=120 possibilities. Even though the seats form a circle geometrically, the numbers distinguish positions.
One can still rotate the seating plan physically, but if labels remain fixed, the assignment to labels changes. The counting equivalence is a statement about outcomes, not about the shape of the venue.
Worked example 24: A clockwise distance is directional
With six people around an unmarked circle, the relative location of Bea two seats clockwise from Alex is different from Bea two seats anticlockwise, unless the problem explicitly treats mirror directions as equivalent.
Fixing Alex, each signed displacement specifies one of the five remaining places. In a directed seating question, clockwise and anticlockwise language must be respected to avoid double counting.
Worked example 25: A pair across a square table
With four distinct guests at four equally spaced circular seats, each guest has one opposite position. Fix A, put B opposite, then arrange the remaining two people in 2!=2 ways.
The total circular count is 3!=6, so the probability that A and B are opposite is 2/6=1/3. The analogous formula for six distinct people gave 1/5.
Worked example 26: Six people with two neighbours fixed
Six distinct people sit in a circle. Fix Alex and require Bea and Chen to occupy the two neighbouring positions, in either order. That gives 2 ways to orient Bea and Chen.
Arrange the remaining three people in 3!=6 ways, producing 12 circular arrangements. Counting each neighbour position as merely “chosen” without recording the two orientations would undercount by two.
Worked example 27: An exact block count with three consecutive names
Among six distinct people, suppose A,B,C must form three consecutive seats in any internal order. Treat the triple as one block alongside the other three people.
There are (4−1)!=6 circular block orders and 3!=6 ways to order the triple, giving 36 arrangements. The triple is one specified group; no separate count of different block starting points is needed.
Worked example 28: A prescribed order inside a three-person block
In the preceding six-person problem, require the three consecutive people to appear specifically A-B-C clockwise. The block now has only one internal orientation.
The number becomes (4−1)!=6, compared with 36 when all internal orders were allowed. The wording “together” differs from “together in this order”, and the factor 3! reflects that distinction.
Worked example 29: The role of symmetry in a proof
The formula (n−1)! is not simply an item to memorise. It follows from a bijection between circular arrangements and linear orders with a chosen anchor.
A strong student should explain why each circular order has exactly one representation after the anchor is fixed, and why no different circular arrangement maps to the same anchored order.
Worked example 30: When an anchor is not unique
With repeated indistinguishable beads, anchoring “a red bead” may be ambiguous because several identical red beads can be chosen. A single circular pattern can be represented by different anchored linear strings.
This is why the distinct-people argument cannot be copied unaltered to necklaces. We need to count orbits of rotations, including patterns that are fixed by some rotations.
Repeated colours, necklaces, bracelets and symmetry proofs
Advanced example 1: Four beads with two red and two blue
On four positions, choose the two red positions in C(4,2)=6 ways. But rotational necklaces contain only two distinct patterns: the reds are adjacent or they lie opposite.
A simple division by four would produce 1.5, which is impossible as a count. Some repeated-colour arrangements remain fixed by a nontrivial rotation, so the number of representations per necklace need not be constant.
Advanced example 2: What Burnside’s principle averages
Burnside’s lemma counts distinct orbits by averaging the number of labelled arrangements fixed under each allowed symmetry. For a four-bead necklace the symmetries are rotations by zero, one, two and three positions.
The identity rotation fixes every colouring. A half-turn may preserve some patterns, while quarter-turns preserve fewer. This fixed-point accounting corrects the failure of dividing by four uniformly.
Advanced example 3: Two red beads under four rotations
For two red and two blue beads on four labelled positions, the identity fixes all six arrangements. The half-turn fixes two opposite-pair colourings; the one-step and three-step rotations fix none.
The orbit count is (6+0+2+0)/4=2. This matches the adjacent-versus-opposite classification and supplies a checkable example of Burnside’s lemma rather than a memorised formula.
Advanced example 4: Four binary beads with no colour-count restriction
Every position can be red or blue, giving sixteen labelled strings. The identity fixes sixteen, quarter-turns each fix the two monochrome strings, and a half-turn fixes four.
The number of rotational necklaces is (16+2+4+2)/4=6. These comprise one all-blue, one all-red, one with one red, one with three red, and two with two red.
Advanced example 5: Five binary beads
There are 32 binary strings on five labelled positions. Because five is prime, a nontrivial rotation cycles through all positions, so it fixes only the two monochrome strings.
The count is (32+4×2)/5=8 necklaces. The calculation explains why dividing 32 by five would be incorrect: some patterns have rotational symmetry and therefore smaller orbits.
Advanced example 6: Six binary beads
For six labelled binary positions, the fixed counts for rotations through zero, one, two, three, four and five positions are 64,2,4,8,4,2.
Their total is 84, giving 84/6=14 rotational necklaces. The values depend on how the rotation partitions positions into cycles; a colouring is fixed only when all positions in each cycle share a colour.
Advanced example 7: Six beads with exactly two red
There are C(6,2)=15 labelled strings with exactly two red beads. The half-turn fixes three, corresponding to the three opposite pairs. The other nontrivial rotations fix none.
Burnside gives (15+3)/6=3 necklaces. Their geometric types are circular red-bead separations of one, two or three positions. This is also a direct classification that checks the orbit calculation.
Advanced example 8: Six beads with three red
There are C(6,3)=20 labelled strings with three red and three blue beads. Rotations by two and four positions each fix two strings; the other nontrivial rotations fix none.
Thus the necklace count is (20+2+2)/6=4. This is a case where different colourings have different rotational stabilisers, making naive division of the labelled count invalid.
Advanced example 9: The alternating necklace
The six-position patterns RBRBRB and BRBRBR are related by a one-position rotation, so they form one rotational necklace type.
If each seat had a permanent number, they would instead be different assignments. The same physical ring can support different counting models depending on which positions are distinguished.
Advanced example 10: Two red beads on five positions
There are C(5,2)=10 labelled placements of two identical reds among five sites. No nontrivial rotation can preserve a colouring with exactly two reds because five is prime.
Every orbit has five members, leaving 10/5=2 necklaces. One type has adjacent reds; the other has a one-position gap along the shorter circular route.
Advanced example 11: Reflecting a necklace to get a bracelet
A necklace usually identifies rotations. A bracelet may also identify mirror images, because the object can be flipped over. For four distinct beads, the six rotational necklaces pair under reflection.
That gives three bracelets. With repeated colours, however, some necklaces may be unchanged under reflection, so dividing every necklace count by two need not work.
Advanced example 12: Six binary bracelets
For six binary beads the rotations have 84 fixed-colouring incidences. Across the six reflections, three axes through opposite beads each fix 16 colourings and three axes through opposite gaps each fix eight.
The combined symmetry group has twelve elements, so the bracelet count is (84+3×16+3×8)/12=13. This optional extension illustrates why reflection fixed points matter.
Advanced example 13: Two-red necklaces versus bracelets
For six positions with two identical red beads and four blue, the three rotational necklace types are characterised by circular separation one, two or three.
Each such separation class is also unchanged as a type under reflection, so the bracelet count remains three. The absence of a reduction is a useful warning that reflection does not always halve a repeated-colour count.
Advanced example 14: Why reflection is not assumed in dining questions
Two diners seated to someone’s immediate left and right have distinct positions in an ordinary circular-table model. Reflecting the entire order swaps left and right and generally changes the relative arrangement.
Thus the standard count (n−1)! ignores rotations but not reflections. A question must explicitly identify mirror images before a further symmetry quotient is appropriate.
Advanced example 15: Uniformly random labelled arrangements
If six distinct people are uniformly assigned to six numbered chairs, there are 6!=720 equally likely seat assignments. Each unnumbered circular order has exactly six numbered representatives.
Therefore the 120 relative orders are also equally likely in this distinct-person model. This argument justifies probability calculations using circular counts rather than merely assuming they are uniform.
Advanced example 16: Randomly coloured necklaces need care
If each of six labelled sites is independently coloured red or blue with equal probability, every labelled string has probability 1/64. But different necklace orbits have different numbers of strings.
Therefore the resulting unlabelled necklace types are not generally equally likely. A probability question about randomly coloured positions must count labelled configurations or weight rotational orbits by their sizes.
Advanced example 17: General adjacency probability
Fix one of n≥3 distinct people at a reference position. The other named person has n−1 equally likely relative positions, exactly two of which are neighbouring.
The adjacency probability is 2/(n−1). Equivalently, the favourable count 2(n−2)! divided by total (n−1)! gives the same result, proving consistency between counting and probability.
Advanced example 18: Nonadjacent pair count
For n≥3 distinct people, there are (n−1)! circular orders in total and 2(n−2)! orders with a given pair adjacent.
Subtracting gives (n−1)!−2(n−2)! nonadjacent orders. When n=3 this is zero because everyone is adjacent to everyone else; checking the smallest case catches errors.
Advanced example 19: Couples as adjacent blocks
Six different people form three named couples. If each couple must sit together, collapse each into a block, arrange three blocks circularly in (3−1)!=2 ways, then orient the three couples in 2³=8 ways.
There are sixteen arrangements, out of 120 total, for probability 2/15 under equally likely circular orders. The block argument must count each seating exactly once.
Advanced example 20: Three names consecutive
Six distinct people sit in a circle; a specified triple must occupy three consecutive seats in any internal order. Treat the triple as one block plus the other three individuals.
The four circular objects have (4−1)!=6 orders, and the triple has 3!=6 internal orders, giving 36 seatings. If one clockwise triple order is prescribed, only six remain.
Advanced example 21: Clockwise order of two names
For five distinct diners, fix A. Of the 4!=24 circular orders of the other people, the orders placing B before C clockwise from A and those placing C before B form equal-sized paired sets.
Therefore 12 seatings put B before C clockwise. The pairing proves the half-count without enumerating all 24 possibilities.
Advanced example 22: Alternating gender categories
For three distinct men and three distinct women alternating in a circle, arrange the women in 2!=2 circular orders. Fill the three intervening positions with the men in 3!=6 orders.
There are twelve alternating arrangements, not 24. Choosing men first would be a different counting method for the same outcomes, not an independent extra factor of two.
Advanced example 23: The purpose of a distinguished anchor
A circular order of distinct people has no natural first seat. Fixing one named participant creates one unique linear description of each rotation class.
This is a bijection, which explains (n−1)! transparently. It fails as a direct argument for indistinguishable beads because selecting a named anchor among identical beads may not be possible.
Advanced example 24: A circular arrangement with marked chairs
When the seats are numbered, the number of assignments of n distinct people is n!, even if the chairs physically form a circle.
Rotating the people changes who occupies each numbered seat. Symmetry reduction is allowed only when those rotated assignments are explicitly considered identical outcomes.
Advanced example 25: A basic orbit-size proof
An orbit contains all labelled configurations reached through allowed rotations. With n distinct beads, rotating by a nonzero number of positions changes the arrangement, so every orbit has n members.
With repeated colours, some rotations may fix the arrangement, reducing orbit size. Burnside’s lemma handles this by averaging fixed configurations rather than assuming a uniform orbit size.
A circular-counting decision checklist
- Are the objects distinct or do some share identical colours?
- Are the seats numbered or otherwise physically distinguished?
- Are rotations equivalent, and are reflections also equivalent?
- Does a restriction concern adjacency, opposing seats, gender alternation or a prescribed clockwise order?
- If repeated colours are present, which nontrivial rotations fix an arrangement?
- If a probability is requested, what is sampled uniformly?
Independent practice with exact answers
Try counting five distinct people around an unmarked circle; six people with a specified pair adjacent; six people with that pair nonadjacent; and three distinct couples all sitting together. Then count necklaces of four positions with two red and two blue beads, and all binary necklaces of length four.
The answers are 4!=24, 2×4!=48, 120−48=72, and 2!×2³=16. There are two necklaces with two red beads, and six binary necklaces when any number of each colour is allowed. In each question, state exactly why rotations and any reflections do or do not create new outcomes.
Frequently asked questions
Does the official H3 syllabus name circular permutations?
The 2027 H3 Mathematics 9820 syllabus assumes H2 counting and names the symmetry principle, combinatorial proofs and bijections. Circular permutations and Burnside examples are educational extensions, not separately asserted compulsory syllabus headings.
When is (n−1)! correct?
For n distinct objects around an unmarked circle when rotations are identified and mirror images are not, the number is (n−1)!. If seats are numbered or repeated objects have additional symmetries, reassess the model.
When are mirror images considered identical?
Only when the problem states that reflection is an allowed equivalence. A flip-over bracelet may have this rule; ordinary table seatings generally do not.
Can I assume all necklace types are equally likely?
Not if the random experiment colours labelled sites uniformly. Repeated-colour necklace types can have different orbit sizes and therefore different probabilities.
Official references and Bukit Timah tuition context
See the SEAB 2027 A-Level directory, combinatorial proofs, bijection counting, and symmetry and double-counting.
The unchanged eduKateSG small-group Mathematics reference describes its three-student model near Sixth Avenue MRT, not a verified current JC1 H3 specialist. Enquire directly about available tutor expertise, suitable groups, lesson timing and fees.
Did you know? In circular counting, the first and most important mathematical step is deciding which arrangements are genuinely the same. After that, the factorial often becomes the easiest part.
