26 chapters · formula sheet

CAT QA formulas

Every formula and shortcut, chapter by chapter, with when to use each. Print it or keep it open while you practise.

Practise QA chapters

Ratio, Proportion & Variation

Open on its own

Ratios sit underneath partnership, mixtures, time and work, and most word problems in Arithmetic. Focus on combining ratios through a common term and on recognising direct, inverse and joint variation quickly.

Ratios

FormulaWhen to use
\(a : b = ka : kb\) for any \(k \neq 0\)Scaling a ratio to whole numbers or to match an actual quantity.
Part of \(N\) in ratio \(a : b\): first part \(= \dfrac{a}{a + b} \times N\)Dividing a total amount in a given ratio.
If \(a : b = m : n\) and \(b : c = p : q\), then \(a : b : c = mp : np : nq\)Joining two ratios that share a common term.
Compounded ratio of \(a : b\) and \(c : d\) is \(ac : bd\); duplicate \(a^2 : b^2\); sub-duplicate \(\sqrt{a} : \sqrt{b}\) (\(a, b > 0\))Ratios of products, areas or square roots of quantities.
If \(\dfrac{a}{b} = \dfrac{c}{d} = \dfrac{e}{f} = k\), then \(\dfrac{pa + qc + re}{pb + qd + rf} = k\) (denominator \(\neq 0\))Several equal ratios given together.
Number \(x\) added to both terms of \(a : b\) to make it \(c : d\): \(x = \dfrac{bc - ad}{d - c}\) (\(c \neq d\))"What must be added to each term" questions.

Proportion

FormulaWhen to use
\(a : b = c : d \iff ad = bc\); fourth proportional \(d = \dfrac{bc}{a}\)Checking or completing a proportion.
Mean proportional of \(a\) and \(b\) \(= \sqrt{ab}\) (\(a, b > 0\)); third proportional to \(a, b\) \(= \dfrac{b^2}{a}\)When \(a : x = x : b\) or \(a : b = b : x\).
Componendo-dividendo: if \(\dfrac{a}{b} = \dfrac{c}{d}\), then \(\dfrac{a + b}{a - b} = \dfrac{c + d}{c - d}\) (\(a \neq b\), \(c \neq d\))Simplifying equations of the form \(\dfrac{x + y}{x - y} = k\).

Variation

FormulaWhen to use
Direct: \(y \propto x \iff y = kx\), so \(\dfrac{y_1}{y_2} = \dfrac{x_1}{x_2}\)Both quantities rise or fall in the same proportion.
Inverse: \(y \propto \dfrac{1}{x} \iff xy = k\), so \(x_1 y_1 = x_2 y_2\)One quantity doubles when the other halves, e.g. speed and time for a fixed distance.
Joint: \(y \propto \dfrac{xz}{w} \iff y = \dfrac{kxz}{w}\)One quantity depends on several others at once.
Partly constant, partly varying: \(y = A + Bx\)Costs with a fixed part and a variable part. Use two data points to find \(A\) and \(B\).

Partnership

FormulaWhen to use
Profit shares \(= C_1 T_1 : C_2 T_2 : C_3 T_3\) (capital \(\times\) time)Partners invest different amounts for different periods.
Changing capital: each partner's weight \(= \sum (\text{capital} \times \text{months at that capital})\)A partner adds or withdraws money part-way through the year.
Working partner: deduct salary or commission first, then split the rest by capital \(\times\) timeOne partner is paid for managing the business.

Shortcuts and traps

  • Ratio is not quantity: a ratio alone never gives actual values. You need one real number to fix the multiplier \(k\).
  • Match the common term: before joining \(a : b\) and \(b : c\), make the two \(b\) values equal using their LCM.
  • Adding to both terms: adding the same positive number to both terms moves the ratio towards 1; it never keeps it unchanged unless \(a = b\).
  • Inverse is not negative: "inversely proportional" means the product is constant, not that one falls by the same amount the other rises.
  • Partnership time units: use the same unit (usually months) for every partner before multiplying.

Percentage

Open on its own

Percentages turn up inside almost every Arithmetic question: profit and loss, interest, data interpretation and growth. Learn these conversions and the successive-change rule and most questions become one-line calculations.

Core formulas

FormulaWhen to use
\(x\%\) of \(N = \dfrac{x}{100} \times N\)Any "what is x% of" calculation.
Percentage change \(= \dfrac{\text{New} - \text{Old}}{\text{Old}} \times 100\)Growth, fall, increase or decrease from a base value. The base is always the old value.
New value \(= \text{Old} \times \left(1 \pm \dfrac{r}{100}\right)\)Applying an increase (+) or decrease (−) of \(r\%\) directly.
If A is \(r\%\) more than B, then B is \(\dfrac{r}{100 + r} \times 100\%\) less than A"How much less/more is the other one" questions.
If A is \(r\%\) less than B, then B is \(\dfrac{r}{100 - r} \times 100\%\) more than AThe reverse of the row above.

Successive changes

FormulaWhen to use
Net change for \(a\%\) then \(b\%\): \(\left(a + b + \dfrac{ab}{100}\right)\%\)Two changes one after the other. Use negative signs for decreases.
Equal rise and fall of \(r\%\): net \(= -\dfrac{r^2}{100}\%\)"Increased by r% and then decreased by r%": always a loss.
Price up \(r\%\), consumption must fall by \(\dfrac{r}{100 + r} \times 100\%\) to keep spending the sameExpenditure = price × quantity questions.
Population after \(n\) years: \(P\left(1 + \dfrac{r}{100}\right)^n\); depreciation: \(P\left(1 - \dfrac{r}{100}\right)^n\)Constant yearly growth or decline.

Fraction equivalents

FractionPercentageFractionPercentage
\(\frac{1}{2}\)50%\(\frac{1}{9}\)11.11%
\(\frac{1}{3}\)33.33%\(\frac{1}{11}\)9.09%
\(\frac{1}{4}\)25%\(\frac{1}{12}\)8.33%
\(\frac{1}{6}\)16.67%\(\frac{1}{15}\)6.67%
\(\frac{1}{7}\)14.29%\(\frac{1}{16}\)6.25%
\(\frac{1}{8}\)12.5%\(\frac{1}{20}\)5%

Shortcuts and traps

  • Fractions are faster: a rise of 25% is "multiply by \(\frac{5}{4}\)"; a fall of 20% is "multiply by \(\frac{4}{5}\)". Chains of changes become products of simple fractions.
  • Percentage points are not percent: a rate moving from 10% to 12% is a 2 percentage-point rise but a 20% increase.
  • Watch the base: "x is what percent of y" divides by y; "by what percent is x more than y" also divides by y, never by x.
  • Undoing a change: to reverse a rise of \(r\%\) you need a fall of \(\frac{r}{100 + r} \times 100\%\), not \(r\%\).

Profit & Loss

Open on its own

Profit and loss questions appear regularly in CAT, often combined with percentages, discounts and successive changes. Be clear about the base of each percentage: profit and loss on cost price, discount on marked price.

Core formulas

FormulaWhen to use
Profit \(= \text{SP} - \text{CP}\); Profit% \(= \dfrac{\text{Profit}}{\text{CP}} \times 100\)Basic profit or loss from cost and selling price.
\(\text{SP} = \text{CP} \times \left(1 + \dfrac{p}{100}\right)\); with a loss, \(\text{SP} = \text{CP} \times \left(1 - \dfrac{l}{100}\right)\)Finding the selling price from a given profit or loss percentage.
\(\text{CP} = \text{SP} \times \dfrac{100}{100 + p}\)Working back to cost price from selling price and profit%.
Profit is \(p\%\) of SP \(\Rightarrow\) profit on CP \(= \dfrac{p}{100 - p} \times 100\%\) (\(p < 100\))When the question measures profit on selling price instead of cost.
CP of \(a\) articles = SP of \(b\) articles \(\Rightarrow\) profit% \(= \dfrac{a - b}{b} \times 100\)"Cost of 12 equals selling price of 10" type. A negative answer is a loss.

Marked price and discount

FormulaWhen to use
Discount \(= \text{MP} - \text{SP}\); Discount% \(= \dfrac{\text{Discount}}{\text{MP}} \times 100\)Any discount question. The base is the marked price.
\(\text{SP} = \text{MP} \times \left(1 - \dfrac{d}{100}\right)\)Selling price after a single discount.
\(\dfrac{\text{MP}}{\text{CP}} = \dfrac{100 + p}{100 - d}\)Linking markup, discount and profit in one step.
Mark up \(m\%\) above CP, then discount \(d\%\): profit% \(= m - d - \dfrac{md}{100}\)Shopkeeper marks up and then offers a discount.
Successive discounts \(d_1\%, d_2\%\): single equivalent \(= d_1 + d_2 - \dfrac{d_1 d_2}{100}\)Two discounts one after another.
Three or more: equivalent \(= \left[1 - \left(1 - \frac{d_1}{100}\right)\left(1 - \frac{d_2}{100}\right)\left(1 - \frac{d_3}{100}\right)\right] \times 100\%\)Chains of discounts.
Buy \(n\), get \(m\) free: effective discount \(= \dfrac{m}{n + m} \times 100\%\)Free-item offers, all items at the same price.

Special cases

FormulaWhen to use
Two items sold at the same SP, one at \(x\%\) gain and one at \(x\%\) loss: net loss \(= \dfrac{x^2}{100}\%\) of total CPOnly when both selling prices are equal and the two percentages are equal.
False weight, sold at cost price: gain% \(= \dfrac{\text{True weight} - \text{False weight}}{\text{False weight}} \times 100\)Dealer gives less than he claims, e.g. 800 g for 1 kg gives 25%.
False weight plus profit \(p\%\) on price: gain% \(= \left[\left(1 + \dfrac{p}{100}\right) \times \dfrac{\text{True weight}}{\text{False weight}} - 1\right] \times 100\)Dealer both marks up the price and cheats on weight.

Shortcuts and traps

  • Profit is on CP: unless the question says otherwise, profit and loss percentages are on cost price, never on selling price.
  • Discounts do not add: 20% and 10% successive discounts equal 28%, not 30%.
  • Equal SP, equal percent: a gain and a loss of the same percentage on equal selling prices is always a net loss.
  • False weight base: divide by the false weight (what the buyer actually gets), not the true weight.
  • Use multipliers: markup 25% then discount 20% is \(\frac{5}{4} \times \frac{4}{5} = 1\), so no profit and no loss.

Simple & Compound Interest

Open on its own

Interest questions in CAT test compound growth, the CI−SI difference and equal instalments. Treat compound interest as successive percentage change and most questions reduce to one multiplier.

Simple interest

FormulaWhen to use
\(\text{SI} = \dfrac{P \times R \times T}{100}\) (\(R\%\) per annum, \(T\) in years)Interest on the original principal only.
Amount \(A = P\left(1 + \dfrac{RT}{100}\right)\)Total repayable after \(T\) years at simple interest.
Sum becomes \(n\) times in \(T\) years: \(R \times T = 100(n - 1)\)"Doubles in 8 years" type. Doubling gives \(RT = 100\).

Compound interest

FormulaWhen to use
\(A = P\left(1 + \dfrac{R}{100}\right)^n\); \(\text{CI} = A - P\)Interest compounded once a year for \(n\) years.
Compounded \(k\) times a year: \(A = P\left(1 + \dfrac{R}{100k}\right)^{kn}\)Half-yearly (\(k = 2\)) or quarterly (\(k = 4\)) compounding at \(R\%\) per annum.
Effective annual rate for half-yearly compounding \(= \left(R + \dfrac{R^2}{400}\right)\%\)Comparing a half-yearly rate with an annual one.
Different rates each year: \(A = P\left(1 + \dfrac{R_1}{100}\right)\left(1 + \dfrac{R_2}{100}\right)\left(1 + \dfrac{R_3}{100}\right)\)The rate changes from year to year.
Amount \(A_1\) after \(n\) years and \(A_2\) after \(n + 1\) years: \(R = \dfrac{A_2 - A_1}{A_1} \times 100\)Two consecutive amounts given, principal unknown.
Sum becomes \(k\) times in \(T\) years at CI \(\Rightarrow\) \(k^m\) times in \(mT\) years"Doubles in 5 years, so 8 times in 15 years."

CI − SI difference

FormulaWhen to use
2 years: \(\text{CI} - \text{SI} = P\left(\dfrac{R}{100}\right)^2\)Same \(P\) and \(R\), annual compounding, 2 years.
3 years: \(\text{CI} - \text{SI} = P\left(\dfrac{R}{100}\right)^2\left(3 + \dfrac{R}{100}\right)\)Same \(P\) and \(R\), annual compounding, 3 years.

Equal instalments

FormulaWhen to use
CI, \(n\) equal annual instalments \(x\) paid at each year end: \(P = \dfrac{x}{1 + r} + \dfrac{x}{(1 + r)^2} + \dots + \dfrac{x}{(1 + r)^n}\), where \(r = \dfrac{R}{100}\)Loan repaid in equal yearly payments at compound interest.
SI, \(n\) equal annual instalments \(x\): \(x\left[n + \dfrac{R\,n(n - 1)}{200}\right] = P\left(1 + \dfrac{Rn}{100}\right)\)Simple-interest instalments, using the usual convention that each instalment earns SI until the end of the term.

Shortcuts and traps

  • First year is equal: with annual compounding, CI and SI are the same for the first year. The difference starts in year 2.
  • Convert rate and time together: half-yearly at 10% per annum for 2 years means 5% for 4 periods, not 10% for 4 periods.
  • Yearly interest forms a GP: at CI, the interest earned in each successive year grows by the factor \(\left(1 + \frac{R}{100}\right)\).
  • Use known multipliers: \(1.1^2 = 1.21\), \(1.1^3 = 1.331\), \(1.05^2 = 1.1025\), \(1.2^2 = 1.44\).
  • Instalments are present values: discount each payment back to the start; do not just divide the final amount by \(n\).

Average

Open on its own

Averages appear in CAT both as direct questions and inside data interpretation, mixtures and speed problems. Think in terms of totals: most average questions are solved by tracking how the sum changes.

Core formulas

FormulaWhen to use
Average \(= \dfrac{\text{Sum of values}}{\text{Number of values}}\); Sum \(= \text{Average} \times n\)Every average question. Convert to totals first.
Weighted average \(= \dfrac{n_1 a_1 + n_2 a_2 + \dots}{n_1 + n_2 + \dots}\)Combining groups of different sizes with different averages.
Two groups with averages \(a_1 < a_2\) and combined average \(A\): \(\dfrac{n_1}{n_2} = \dfrac{a_2 - A}{A - a_1}\)Finding the ratio of group sizes from averages (the alligation rule).
Terms in arithmetic progression: average \(= \dfrac{\text{first} + \text{last}}{2}\)Consecutive numbers, evenly spaced values.
First \(n\) natural numbers: \(\dfrac{n + 1}{2}\); first \(n\) odd numbers: \(n\); first \(n\) even numbers: \(n + 1\)Standard sequences.
Squares \(1^2, \dots, n^2\): \(\dfrac{(n + 1)(2n + 1)}{6}\); cubes \(1^3, \dots, n^3\): \(\dfrac{n(n + 1)^2}{4}\)Average of the first \(n\) squares or cubes.

Changes to the group

FormulaWhen to use
Add \(k\) to every value: average rises by \(k\); multiply every value by \(k\): average is multiplied by \(k\)Uniform changes such as a bonus to everyone or a scaling of marks.
\(n\) members, average \(A\); one joins and average becomes \(A + d\): newcomer \(= A + (n + 1)d\)A new member changes the average. Use \(d < 0\) if the average falls.
\(n\) members, average \(A\); one leaves and average of the rest is \(A + d\): leaver \(= A - (n - 1)d\)A member drops out.
One member replaced, average of \(n\) changes by \(d\): new member \(-\) old member \(= nd\)Replacement questions, e.g. a player substituted in a team.
Value wrongly recorded: correct average \(= \text{wrong average} + \dfrac{\text{correct value} - \text{wrong value}}{n}\)Correcting a misread or miscopied entry.

Average speed

FormulaWhen to use
Average speed \(= \dfrac{\text{Total distance}}{\text{Total time}}\)Always valid. Use when the other rows do not fit.
Equal distances at \(u\) and \(v\): \(\dfrac{2uv}{u + v}\)Onward and return journeys over the same distance.
Equal times at \(u\) and \(v\): \(\dfrac{u + v}{2}\)Same duration spent at each speed.
Three equal distances at \(u, v, w\): \(\dfrac{3uvw}{uv + vw + wu}\)Three legs of equal length.

Shortcuts and traps

  • Deviation method: assume a convenient average, add up the deviations of each value from it, and divide by \(n\) to correct.
  • Averages of averages: you cannot average two group averages directly unless the groups are the same size. Weight them.
  • Average speed: it is the simple mean of speeds only when equal time is spent at each. For equal distances it is lower.
  • Ages move together: the average age of an unchanged group rises by exactly 1 each year.
  • Combined average lies between: a combined average is always between the smallest and largest group averages. Use this to eliminate options.

Mixture, Alligation & Replacement

Open on its own

Mixture questions in CAT are weighted averages in disguise, and replacement questions are repeated percentage decrease. Learn the alligation rule and the replacement formula, and always track the quantity of one component.

Mixing and alligation

FormulaWhen to use
Mean of mixture \(M = \dfrac{q_1 c_1 + q_2 c_2}{q_1 + q_2}\)Price, concentration or strength of a mixture of two quantities.
Alligation: \(\dfrac{q_1}{q_2} = \dfrac{c_2 - M}{M - c_1}\) where \(c_1 < M < c_2\)Finding the ratio in which two items must be mixed to get a mean value.
Sum lent partly at \(r_1\%\) and partly at \(r_2\%\) simple interest, overall \(r\%\) for the same time: \(\dfrac{P_1}{P_2} = \dfrac{r_2 - r}{r - r_1}\)Alligation applied to interest rates, with principal as the weight.
Fraction of A in a mixture with ratio \(A : B = a : b\) is \(\dfrac{a}{a + b}\)Turn each ratio into a fraction of one component before alligating two mixtures.
Mixing several mixtures: quantity of each component in the result \(= \) sum of that component across all mixturesCombining three or more mixtures, or mixtures in unequal amounts.
Mixture sold at SP with \(p\%\) profit: CP of mixture \(= \text{SP} \times \dfrac{100}{100 + p}\), then alligate on CPMixing grades of a product to sell at a target profit.

Dilution and concentration

FormulaWhen to use
Quantity of solute stays fixed: \(Q_1 c_1 = Q_2 c_2\)Adding water or evaporating water, with only the solvent changing.
Water \(W\) added to quantity \(Q\) to bring concentration from \(c_1\) to \(c_2\): \(W = \dfrac{Q(c_1 - c_2)}{c_2}\)Diluting a solution to a target strength.

Removal and replacement

FormulaWhen to use
Vessel of volume \(V\), remove \(x\) and refill with water, \(n\) times: original liquid left \(= V\left(1 - \dfrac{x}{V}\right)^n\)Repeated draw-and-refill with the volume kept constant.
Concentration after \(n\) replacements: \(c_n = c_0\left(1 - \dfrac{x}{V}\right)^n\)Starting from a solution of strength \(c_0\) instead of a pure liquid.
Different amounts each time: \(c_n = c_0\left(1 - \dfrac{x_1}{V}\right)\left(1 - \dfrac{x_2}{V}\right) \cdots \left(1 - \dfrac{x_n}{V}\right)\)The amount removed changes between operations, volume still constant.
Refilled with a solution of strength \(c_r\) instead of water: \(c_n = c_r + (c_0 - c_r)\left(1 - \dfrac{x}{V}\right)^n\)Each refill adds some of the same component back.

Shortcuts and traps

  • Cross the right way: the quantity of the cheaper item goes with (dearer − mean), and the dearer item goes with (mean − cheaper).
  • Same base for alligation: both values must be per the same unit, e.g. price per kg or fraction of milk. Never alligate a ratio like 3 : 2 directly.
  • Replacement needs a constant volume: the formula \(\left(1 - \frac{x}{V}\right)^n\) only works if exactly what is removed is put back each time.
  • Track one component: follow the quantity of milk (or acid, or salt) through each step; the other component is just the remainder.

Time & Work

Open on its own

Time and work questions in CAT test combined rates, efficiency comparisons, pipes and cisterns and sharing of wages. Work with rates (work per day), not times, and add rates when people or pipes work together.

Core formulas

FormulaWhen to use
A finishes a job in \(a\) days \(\Rightarrow\) A's rate \(= \dfrac{1}{a}\) of the job per dayConverting a time into a rate before combining.
A and B together: time \(= \dfrac{ab}{a + b}\)Two people working together at constant rates.
A, B and C together: time \(= \dfrac{abc}{ab + bc + ca}\)Three people working together.
Together in \(T\) days, A alone in \(a\) days \(\Rightarrow\) B alone \(= \dfrac{aT}{a - T}\) (\(a > T\))Finding one person's time from the joint time.
A works \(d\) days and leaves; B finishes the rest in \(\left(1 - \dfrac{d}{a}\right) \times b\) daysOne person starts the job and another completes it.
\(\dfrac{M_1 D_1 H_1}{W_1} = \dfrac{M_2 D_2 H_2}{W_2}\)Men, days, hours and amount of work change together (equal efficiency per person).
A is \(k\) times as efficient as B \(\Rightarrow\) A takes \(\dfrac{1}{k}\) of B's timeEfficiency and time are inversely proportional for a fixed job.

Pipes and cisterns

FormulaWhen to use
Inlet fills in \(a\) hours: rate \(+\dfrac{1}{a}\); outlet empties in \(b\) hours: rate \(-\dfrac{1}{b}\)Treat emptying as negative work.
Inlet (\(a\) h) and outlet (\(b\) h) open together, \(b > a\): tank fills in \(\dfrac{ab}{b - a}\) hoursOne pipe fills while another empties.
Pipe fills in \(a\) h alone but \(t\) h with a leak (\(t > a\)): leak empties a full tank in \(\dfrac{at}{t - a}\) hoursFinding the strength of a leak.

Work and wages

FormulaWhen to use
Share of wages \(= \dfrac{\text{work done by the person}}{\text{total work}} \times \text{total wage}\)Payment is for the job, split by work contributed.
Work done \(= \text{rate} \times \text{days worked}\); if all work the same days, shares are in the ratio of rates \(\dfrac{1}{a} : \dfrac{1}{b} : \dfrac{1}{c}\)Splitting wages when people work together for the whole job.

Shortcuts and traps

  • Add rates, not times: 10 days and 15 days together is 6 days, not 25 or 12.5.
  • LCM as total work: take the job as the LCM of the individual times, e.g. 30 units for 10 and 15 days, so rates are 3 and 2 units per day.
  • Alternate days: find the work done per full cycle, then check whether the job ends part-way through the last cycle.
  • Wages follow work: a faster worker earns more for the same days; pay is not split by days alone.
  • Common unit for mixed workers: convert men, women and children into one unit (e.g. 2 men = 3 women) before applying the man-days formula.

Time, Speed & Distance

Open on its own

Time, Speed and Distance gives CAT one or two questions almost every year, usually dressed up as trains, boats, races or runners on a circular track. Focus on average speed, relative speed and the inverse link between speed and time.

Core formulas

FormulaWhen to use
\(\text{Distance} = \text{Speed} \times \text{Time}\)The base relation behind every question in this chapter.
\(1 \text{ km/h} = \dfrac{5}{18} \text{ m/s}\); \(1 \text{ m/s} = \dfrac{18}{5} \text{ km/h}\)Converting units, especially when train lengths are in metres.
Average speed \(= \dfrac{\text{Total distance}}{\text{Total time}}\)Any journey with several legs. Never average the speeds directly.
Equal distances at speeds \(a\) and \(b\): average speed \(= \dfrac{2ab}{a + b}\)Going and returning over the same route at different speeds.
Equal times at speeds \(a\) and \(b\): average speed \(= \dfrac{a + b}{2}\)Travelling for the same length of time at each speed.
For a fixed distance, if speeds are in ratio \(a : b\), times are in ratio \(b : a\)Comparing two people or vehicles covering the same distance.

Relative speed and trains

FormulaWhen to use
Relative speed \(= a + b\) (opposite directions), \(|a - b|\) (same direction)Two bodies moving towards, away from, or after each other.
Train of length \(L\) at speed \(s\) passes a pole or a person standing still in \(\dfrac{L}{s}\)Crossing an object whose length can be ignored.
Train of length \(L\) passes a platform of length \(P\) in \(\dfrac{L + P}{s}\)Crossing a platform, bridge or tunnel.
Two trains of lengths \(L_1, L_2\) cross each other in \(\dfrac{L_1 + L_2}{s_1 + s_2}\) (opposite) or \(\dfrac{L_1 + L_2}{|s_1 - s_2|}\) (same direction)Two moving trains passing each other completely.
A and B start together from opposite ends towards each other; after meeting they take \(t_1\) and \(t_2\) to finish: \(\dfrac{\text{Speed of A}}{\text{Speed of B}} = \sqrt{\dfrac{t_2}{t_1}}\)Questions that give only the times taken after the meeting point.

Boats, races and circular tracks

FormulaWhen to use
Boat speed \(u\) in still water, stream speed \(v\) (\(u > v\)): downstream \(= u + v\), upstream \(= u - v\)Any boat, swimmer or flight with wind or current.
\(u = \dfrac{\text{Down} + \text{Up}}{2}\), \(v = \dfrac{\text{Down} - \text{Up}}{2}\)Finding still-water and stream speeds from downstream and upstream speeds.
A beats B by \(x\) m in an \(L\) m race: \(\dfrac{\text{Speed of A}}{\text{Speed of B}} = \dfrac{L}{L - x}\)When A finishes, B has run \(L - x\) m in the same time.
Track of length \(L\), speeds \(a > b\), same start: first meeting after \(\dfrac{L}{a + b}\) (opposite) or \(\dfrac{L}{a - b}\) (same direction)Runners on a circular track meeting anywhere on it.
First meeting at the starting point after \(\text{LCM}\left(\dfrac{L}{a}, \dfrac{L}{b}\right)\)When both must be back at the start together. Works for any direction.
Speeds in ratio \(p : q\) in lowest terms: number of distinct meeting points \(= |p - q|\) (same direction), \(p + q\) (opposite)"At how many different points on the track do they meet" questions.

Shortcuts and traps

  • Average speed is not the average of speeds: 40 km/h one way and 60 km/h back gives 48 km/h, not 50 km/h.
  • Check the units first: lengths in metres and speeds in km/h must be converted with \(\frac{5}{18}\) before dividing.
  • Train lengths always add: when two trains cross, the distance is the sum of both lengths, whatever the direction.
  • Beating by time: if A beats B by \(x\) m and, equivalently, by \(t\) s, then B runs those \(x\) m in \(t\) s, so B's speed is \(\frac{x}{t}\).
  • Reduce the ratio: the meeting-point count uses the speed ratio in lowest terms; 6 : 4 must become 3 : 2 first.

Linear Equations

Open on its own

Linear equations sit inside most CAT word problems, and some questions test only whether a system has one, none or many solutions. Learn the consistency conditions and how to list integer solutions of \(ax + by = c\).

Two variables

FormulaWhen to use
System \(a_1x + b_1y = c_1\), \(a_2x + b_2y = c_2\) has a unique solution if \(a_1b_2 - a_2b_1 \neq 0\), i.e. \(\dfrac{a_1}{a_2} \neq \dfrac{b_1}{b_2}\)Checking whether two lines intersect at exactly one point.
No solution if \(\dfrac{a_1}{a_2} = \dfrac{b_1}{b_2} \neq \dfrac{c_1}{c_2}\)Parallel lines. Ratios assume non-zero denominators.
Infinitely many solutions if \(\dfrac{a_1}{a_2} = \dfrac{b_1}{b_2} = \dfrac{c_1}{c_2}\)The two equations are the same line.
\(x = \dfrac{c_1b_2 - c_2b_1}{a_1b_2 - a_2b_1}\), \(y = \dfrac{a_1c_2 - a_2c_1}{a_1b_2 - a_2b_1}\)Direct solution (Cramer's rule) when \(a_1b_2 - a_2b_1 \neq 0\).
\(a_1x + b_1y = 0\), \(a_2x + b_2y = 0\) has a non-zero solution only if \(a_1b_2 = a_2b_1\)Homogeneous systems; otherwise \(x = y = 0\) is the only solution.

Three or more variables

FormulaWhen to use
\(n\) unknowns need \(n\) independent equations for a unique solutionDeciding quickly whether a system can be solved completely.
Three equations have a unique solution if the determinant of coefficients \(a_1(b_2c_3 - b_3c_2) - b_1(a_2c_3 - a_3c_2) + c_1(a_2b_3 - a_3b_2) \neq 0\)Systems in \(x, y, z\) with coefficients \(a_i, b_i, c_i\).
Fewer equations than unknowns: find \(p, q\) so that \(p \times (\text{eq. 1}) + q \times (\text{eq. 2})\) matches the asked expression"Find \(x + y + z\)" when the individual values cannot be found.
\(x + y = a\), \(y + z = b\), \(z + x = c\): \(x + y + z = \dfrac{a + b + c}{2}\), \(x = \dfrac{a - b + c}{2}\)Symmetric pairwise sums; add all equations first.

Integer solutions

FormulaWhen to use
\(ax + by = c\) (integers \(a, b\) not both zero) has integer solutions only if \(\gcd(a, b)\) divides \(c\)Checking first whether any whole-number answer exists.
If \((x_0, y_0)\) is one solution and \(g = \gcd(a, b)\): all solutions are \(x = x_0 + \dfrac{b}{g}t\), \(y = y_0 - \dfrac{a}{g}t\), \(t\) an integerListing every solution, then applying limits such as \(x, y > 0\).
For \(a, b, c > 0\) and \(\gcd(a, b) = 1\), the number of non-negative solutions of \(ax + by = c\) is \(\left\lfloor \dfrac{c}{ab} \right\rfloor\) or \(\left\lfloor \dfrac{c}{ab} \right\rfloor + 1\)Quick estimate of the count; confirm by listing.
\(x_1 + x_2 + \dots + x_r = n\): non-negative solutions \(= \binom{n + r - 1}{r - 1}\), positive solutions \(= \binom{n - 1}{r - 1}\)Counting whole-number solutions when all coefficients are 1.

Shortcuts and traps

  • Step size: in \(ax + by = c\) with \(\gcd(a, b) = 1\), consecutive solutions move \(x\) by \(b\) and \(y\) by \(a\) in opposite directions. Find one solution, then step.
  • Positive versus non-negative: for positive solutions, put \(x = x' + 1\), \(y = y' + 1\) and count non-negative solutions of \(ax' + by' = c - a - b\).
  • Use divisibility: in \(3x + 5y = 47\), \(47 - 5y\) must be a multiple of 3, which fixes \(y\) modulo 3 at once.
  • Zero coefficients: the ratio tests fail when a coefficient in the second equation is 0; use \(a_1b_2 - a_2b_1\) instead.

Quadratic Equations

Open on its own

Quadratic equations appear in CAT both directly and inside questions on sign of roots, ranges of a parameter and maximum or minimum values. Focus on the discriminant, sum and product of roots, and the vertex of the parabola.

Roots and discriminant

FormulaWhen to use
\(ax^2 + bx + c = 0\), \(a \neq 0\): \(x = \dfrac{-b \pm \sqrt{b^2 - 4ac}}{2a}\)Finding roots when the equation does not factorise easily.
Discriminant \(D = b^2 - 4ac\): \(D > 0\) two distinct real roots, \(D = 0\) equal real roots, \(D < 0\) no real rootsNature of roots; "real roots" means \(D \ge 0\).
\(a, b, c\) rational and \(D\) a perfect square of a rational: roots are rationalChecking whether roots are rational or integers.
Rational \(a, b, c\): if \(p + \sqrt{q}\) is a root (\(p, q\) rational, \(\sqrt{q}\) irrational), so is \(p - \sqrt{q}\)Irrational roots come in conjugate pairs.

Sum, product and forming equations

FormulaWhen to use
\(\alpha + \beta = -\dfrac{b}{a}\), \(\alpha\beta = \dfrac{c}{a}\)Any expression in the roots without solving the equation.
\(|\alpha - \beta| = \dfrac{\sqrt{D}}{|a|}\)Difference of roots, for \(D \ge 0\).
\(\alpha^2 + \beta^2 = (\alpha + \beta)^2 - 2\alpha\beta\); \(\alpha^3 + \beta^3 = (\alpha + \beta)^3 - 3\alpha\beta(\alpha + \beta)\)Symmetric expressions in the roots.
Equation with roots \(\alpha, \beta\): \(x^2 - (\alpha + \beta)x + \alpha\beta = 0\)Forming a quadratic from its sum and product of roots.
Roots \(\dfrac{1}{\alpha}, \dfrac{1}{\beta}\): \(cx^2 + bx + a = 0\) (\(c \neq 0\)); roots \(-\alpha, -\beta\): \(ax^2 - bx + c = 0\); roots \(k\alpha, k\beta\): \(ax^2 + kbx + k^2c = 0\)New equation whose roots are changed versions of the old roots.
Cubic \(ax^3 + bx^2 + cx + d = 0\): \(\alpha + \beta + \gamma = -\dfrac{b}{a}\), \(\alpha\beta + \beta\gamma + \gamma\alpha = \dfrac{c}{a}\), \(\alpha\beta\gamma = -\dfrac{d}{a}\)Sum and product questions on cubic equations.

Sign of roots and extreme values

FormulaWhen to use
Both roots positive: \(D \ge 0\), \(-\dfrac{b}{a} > 0\), \(\dfrac{c}{a} > 0\)Range of a parameter for two positive roots.
Both roots negative: \(D \ge 0\), \(-\dfrac{b}{a} < 0\), \(\dfrac{c}{a} > 0\)Range of a parameter for two negative roots.
Roots of opposite signs: \(\dfrac{c}{a} < 0\)\(D > 0\) follows automatically in this case.
One root zero: \(c = 0\); roots equal in size, opposite in sign: \(b = 0\) and \(\dfrac{c}{a} < 0\)Special root conditions.
\(y = ax^2 + bx + c\) has its vertex at \(x = -\dfrac{b}{2a}\), extreme value \(-\dfrac{D}{4a}\): minimum if \(a > 0\), maximum if \(a < 0\)Maximum or minimum of a quadratic expression over all real \(x\).
\(ax^2 + bx + c > 0\) for all real \(x\) if and only if \(a > 0\) and \(D < 0\)"Always positive" questions. For "always negative", \(a < 0\) and \(D < 0\).

Shortcuts and traps

  • Check the leading coefficient: if \(a\) contains a parameter, the value that makes \(a = 0\) turns the equation linear and must be handled separately.
  • Sign conditions need real roots: sum and product alone are not enough; also require \(D \ge 0\).
  • Do not cancel the variable: dividing \(x^2 = 3x\) by \(x\) loses the root \(x = 0\).
  • Quick root test: if \(a + b + c = 0\), then \(x = 1\) is a root and the other root is \(\frac{c}{a}\).

Surds & Indices

Open on its own

Surds and indices are tested in CAT through simplification, equations with powers and comparison of roots, and they also support questions in other algebra chapters. Focus on the laws of exponents, rationalising, and bringing surds to a common root before comparing.

Laws of indices

FormulaWhen to use
\(a^m \times a^n = a^{m + n}\), \(\dfrac{a^m}{a^n} = a^{m - n}\)Multiplying or dividing powers of the same base (\(a > 0\) for non-integer powers).
\((a^m)^n = a^{mn}\)Power of a power.
\((ab)^n = a^n b^n\), \(\left(\dfrac{a}{b}\right)^n = \dfrac{a^n}{b^n}\)Same power on different bases (\(a, b > 0\) for non-integer \(n\), \(b \neq 0\)).
\(a^0 = 1\) and \(a^{-n} = \dfrac{1}{a^n}\), for \(a \neq 0\)Zero and negative exponents.
\(a^{p/q} = \sqrt[q]{a^p} = \left(\sqrt[q]{a}\right)^p\), for \(a > 0\) and integers \(p, q\) with \(q > 0\)Converting between fractional powers and roots.
\(a^x = a^y \Rightarrow x = y\), for \(a > 0\), \(a \neq 1\)Solving exponential equations by matching bases.
\(x^n = y^n \Rightarrow x = y\), for \(x, y > 0\) and \(n \neq 0\)Solving equations by matching exponents.

Surds and rationalising

FormulaWhen to use
\(\sqrt{a} \times \sqrt{b} = \sqrt{ab}\), \(\dfrac{\sqrt{a}}{\sqrt{b}} = \sqrt{\dfrac{a}{b}}\), for \(a \ge 0\), \(b > 0\)Combining or splitting square roots.
\(\dfrac{1}{\sqrt{a} + \sqrt{b}} = \dfrac{\sqrt{a} - \sqrt{b}}{a - b}\), for \(a \neq b\)Removing a surd from the denominator.
\(\dfrac{1}{a + \sqrt{b}} = \dfrac{a - \sqrt{b}}{a^2 - b}\), for \(a^2 \neq b\)Rationalising a rational number plus a surd.
\(\sqrt{a + 2\sqrt{b}} = \sqrt{x} + \sqrt{y}\) and \(\sqrt{a - 2\sqrt{b}} = \sqrt{x} - \sqrt{y}\), where \(x + y = a\), \(xy = b\), \(x \ge y \ge 0\)Square root of a surd expression. Example: \(\sqrt{5 + 2\sqrt{6}} = \sqrt{3} + \sqrt{2}\).
\(\sqrt{a + \sqrt{b}} = \sqrt{\dfrac{a + c}{2}} + \sqrt{\dfrac{a - c}{2}}\), where \(c = \sqrt{a^2 - b}\), for \(a > 0\), \(b \ge 0\), \(a^2 \ge b\)Same idea when the inner term is not written as \(2\sqrt{b}\). It simplifies only when \(a^2 - b\) is a perfect square.
\(\sqrt{x + \sqrt{x + \sqrt{x + \cdots}}} = \dfrac{1 + \sqrt{1 + 4x}}{2}\); \(\sqrt{x\sqrt{x\sqrt{x \cdots}}} = x\), for \(x > 0\)Infinite nested roots: set the expression equal to \(y\) and square.

Comparing surds

FormulaWhen to use
To compare \(\sqrt[m]{a}\) and \(\sqrt[n]{b}\) (\(a, b > 0\)), raise both to the power \(\text{LCM}(m, n)\)Roots of different orders. Example: \(\sqrt{2}\) vs \(\sqrt[3]{3}\) becomes \(8\) vs \(9\).
\(\sqrt{a + 1} - \sqrt{a} = \dfrac{1}{\sqrt{a + 1} + \sqrt{a}}\), which decreases as \(a\) increases (\(a \ge 0\))Comparing differences such as \(\sqrt{12} - \sqrt{11}\) and \(\sqrt{11} - \sqrt{10}\).
For \(x > 1\), \(x^m\) increases with \(m\); for \(0 < x < 1\), \(x^m\) decreases as \(m\) increasesOrdering powers of the same base.
\(a + \sqrt{b} = c + \sqrt{d}\) with \(a, b, c, d\) rational and \(\sqrt{b}\), \(\sqrt{d}\) irrational gives \(a = c\), \(b = d\)Equating rational and irrational parts.

Shortcuts and traps

  • Roots do not split over sums: \(\sqrt{a + b} \neq \sqrt{a} + \sqrt{b}\) in general; \(\sqrt{9 + 16} = 5\), not 7.
  • Tower powers go top-down: \(2^{3^2} = 2^9\), while \((2^3)^2 = 2^6\).
  • Square root of a square: \(\sqrt{x^2} = |x|\), not \(x\), when \(x\) can be negative.
  • Base 1 and base 0: \(a^x = a^y\) gives no information when \(a = 1\), and \(0^0\) is undefined. Check these cases separately in equations.

Logarithms

Open on its own

Logarithms give CAT one question in most years, usually simplification, an equation or an inequality where the domain decides the answer. Learn the laws, change of base, and always check that every log is defined.

Definition and laws

FormulaWhen to use
\(\log_a x = y \iff a^y = x\), for \(a > 0\), \(a \neq 1\), \(x > 0\)Converting between log form and exponent form.
\(\log_a (mn) = \log_a m + \log_a n\); \(\log_a \dfrac{m}{n} = \log_a m - \log_a n\)Splitting or combining logs of products and quotients (\(m, n > 0\)).
\(\log_a m^p = p \log_a m\), for \(m > 0\)Bringing a power down in front of the log.
\(\log_a a = 1\), \(\log_a 1 = 0\)Simplifying terms where the argument equals the base or 1.
\(a^{\log_a x} = x\); \(a^{\log_b c} = c^{\log_b a}\), for \(a, c > 0\)Expressions with a log in the exponent.

Change of base

FormulaWhen to use
\(\log_a b = \dfrac{\log_c b}{\log_c a}\), for any valid base \(c\)Logs with different bases; usually convert to base 10.
\(\log_a b \times \log_b a = 1\), i.e. \(\log_a b = \dfrac{1}{\log_b a}\), for \(b \neq 1\)Bases and arguments swapped.
\(\log_a b \times \log_b c = \log_a c\)Chains of logs that cancel like fractions.
\(\log_{a^q} b^p = \dfrac{p}{q} \log_a b\), for \(q \neq 0\)Base and argument are both powers.

Domain, sign and characteristic

FormulaWhen to use
\(\log_{g(x)} f(x)\) is defined only if \(f(x) > 0\), \(g(x) > 0\), \(g(x) \neq 1\)Finding the domain and rejecting false roots.
If \(a > 1\): \(\log_a x > \log_a y \iff x > y > 0\). If \(0 < a < 1\) the inequality reversesLog inequalities.
\(\log_a x > 0\) when \(a\) and \(x\) are on the same side of 1; \(\log_a x < 0\) when they are on opposite sides (\(x > 0\))Sign of a log without calculating it.
Number of digits in an integer \(N \ge 1\) \(= \lfloor \log_{10} N \rfloor + 1\)"How many digits does \(2^{100}\) have" questions.
For \(0 < N < 1\), zeros between the decimal point and the first non-zero digit \(= -\lfloor \log_{10} N \rfloor - 1\)Leading zeros in small decimals such as \(\left(\frac{1}{2}\right)^{20}\).

Common log values

Number\(\log_{10}\)Number\(\log_{10}\)
20.301060.7782
30.477170.8451
40.602180.9031
50.699090.9542

Shortcuts and traps

  • No law for sums: \(\log (m + n) \neq \log m + \log n\). Only products, quotients and powers split.
  • Check the domain last: after solving, substitute each root back and reject any that make an argument or base invalid.
  • Even powers: \(\log_a x^2 = 2 \log_a |x|\), not \(2 \log_a x\), because \(x\) may be negative.
  • Base below 1 flips inequalities: \(\log_{0.5} x > \log_{0.5} y\) means \(x < y\).
  • Build other values: \(\log 5 = 1 - \log 2\) and \(\log 6 = \log 2 + \log 3\), so only \(\log 2\), \(\log 3\) and \(\log 7\) need memorising.

Progressions

Open on its own

Progressions give CAT one or two questions a year, often as a term or sum problem, and the AM-GM inequality is used for maximum and minimum questions across algebra. Focus on the nth term and sum formulas, infinite GP, and the standard sums of naturals.

Arithmetic progression (AP)

FormulaWhen to use
\(T_n = a + (n - 1)d\)Any term of an AP with first term \(a\) and common difference \(d\).
\(S_n = \dfrac{n}{2}\left[2a + (n - 1)d\right] = \dfrac{n}{2}(a + l)\)Sum of \(n\) terms; use the second form when the last term \(l\) is known.
Number of terms from \(a\) to \(l\) \(= \dfrac{l - a}{d} + 1\)Counting terms, for example multiples of 7 between two numbers.
\(a, b, c\) in AP \(\iff 2b = a + c\); take three terms as \(a - d, a, a + d\)Problems that give the sum of three terms in AP.
\(T_n = S_n - S_{n-1}\) for \(n \ge 2\), and \(T_1 = S_1\)Finding a term when the sum of \(n\) terms is given as a formula.

Geometric progression (GP)

FormulaWhen to use
\(T_n = ar^{n-1}\)Any term of a GP with first term \(a\) and common ratio \(r\).
\(S_n = \dfrac{a(r^n - 1)}{r - 1}\) for \(r \neq 1\); \(S_n = na\) for \(r = 1\)Sum of \(n\) terms of a GP.
\(S_\infty = \dfrac{a}{1 - r}\), for \(|r| < 1\)Sum of an infinite GP; it does not exist for \(|r| \ge 1\) (with \(a \neq 0\)).
\(a, b, c\) in GP \(\iff b^2 = ac\); take three terms as \(\dfrac{a}{r}, a, ar\)Problems that give the product of three terms in GP.
Product of \(n\) positive terms of a GP \(= (\text{first} \times \text{last})^{n/2}\)Product questions without finding every term.

HP and means

FormulaWhen to use
\(a, b, c\) in HP \(\iff \dfrac{1}{a}, \dfrac{1}{b}, \dfrac{1}{c}\) in AP; then \(b = \dfrac{2ac}{a + c}\)Harmonic progressions: work with reciprocals as an AP (all terms non-zero).
For \(a, b > 0\): AM \(= \dfrac{a + b}{2}\), GM \(= \sqrt{ab}\), HM \(= \dfrac{2ab}{a + b}\), and AM \(\ge\) GM \(\ge\) HMComparing means. Equality holds only when \(a = b\).
For \(a, b > 0\): \(\text{GM}^2 = \text{AM} \times \text{HM}\)Finding one mean of two numbers from the other two.
For \(x_1, \dots, x_n > 0\): \(\dfrac{x_1 + \dots + x_n}{n} \ge \sqrt[n]{x_1 x_2 \cdots x_n}\), equality when all are equalMinimum of a sum when the product is fixed, or maximum of a product when the sum is fixed.

Standard sums

FormulaWhen to use
\(1 + 2 + \dots + n = \dfrac{n(n + 1)}{2}\)Sum of the first \(n\) natural numbers.
\(1^2 + 2^2 + \dots + n^2 = \dfrac{n(n + 1)(2n + 1)}{6}\)Sum of squares, for example counting squares on a chessboard.
\(1^3 + 2^3 + \dots + n^3 = \left[\dfrac{n(n + 1)}{2}\right]^2\)Sum of cubes; it equals the square of the sum of naturals.
First \(n\) odd numbers sum to \(n^2\); first \(n\) even numbers sum to \(n(n + 1)\)Quick sums of odd or even numbers.

Shortcuts and traps

  • Sum formula tells you the type: if \(S_n = pn^2 + qn\) with no constant term, the series is an AP with \(d = 2p\). A constant term means the first term breaks the pattern.
  • Four terms in AP: take \(a - 3d, a - d, a + d, a + 3d\), whose common difference is \(2d\).
  • Infinite GP needs \(|r| < 1\): check the ratio before using \(\frac{a}{1 - r}\).
  • AM-GM needs positive numbers: the inequality can fail if any term is negative, and the bound is reached only if all terms can be equal.

Functions & Graphs

Open on its own

Functions questions in CAT test domain and range, composition, graph shifts and the max, min and floor functions, often inside a single tricky expression. Sketch the graph whenever you can; most of these questions are faster drawn than solved.

Domain, range and symmetry

FormulaWhen to use
\(\sqrt{g(x)}\) needs \(g(x) \ge 0\); \(\dfrac{1}{g(x)}\) needs \(g(x) \neq 0\); \(\log g(x)\) needs \(g(x) \gt 0\)Finding the domain of a combined expression: intersect all the conditions.
Even: \(f(-x) = f(x)\), graph symmetric about the y-axis. Odd: \(f(-x) = -f(x)\), graph symmetric about the originChecking symmetry or simplifying sums like \(f(a) + f(-a)\).
Even \(\times\) even and odd \(\times\) odd are even; even \(\times\) odd is oddDeciding the parity of a product without expanding it.
An odd function defined at \(x = 0\) has \(f(0) = 0\)Quick check or a free value in functional-equation questions.

Composition and inverse

FormulaWhen to use
\((f \circ g)(x) = f(g(x))\), defined where \(x\) is in the domain of \(g\) and \(g(x)\) is in the domain of \(f\)Nested functions. In general \(f(g(x)) \neq g(f(x))\).
\(f^{-1}\) exists only if \(f\) is one-one; then \(f(f^{-1}(x)) = x\) on the range of \(f\) and \(f^{-1}(f(x)) = x\) on the domain of \(f\)Finding or using an inverse: swap \(x\) and \(y\), then solve for \(y\).
The graph of \(f^{-1}\) is the reflection of the graph of \(f\) in the line \(y = x\)Graph-matching questions and intersection points of \(f\) and \(f^{-1}\).
If \(f(x) = \dfrac{ax + b}{cx + d}\) with \(ad - bc \neq 0\), then \(f^{-1}(x) = \dfrac{dx - b}{a - cx}\)Inverting a linear fraction in one step.

Transformations of graphs (k > 0)

FormulaWhen to use
\(y = f(x) + k\) shifts the graph up by \(k\); \(y = f(x) - k\) shifts it down by \(k\)Vertical shifts.
\(y = f(x - k)\) shifts the graph right by \(k\); \(y = f(x + k)\) shifts it left by \(k\)Horizontal shifts. The sign works opposite to what it looks like.
\(y = -f(x)\) reflects in the x-axis; \(y = f(-x)\) reflects in the y-axisReflections.
\(y = |f(x)|\): parts below the x-axis flip above it. \(y = f(|x|)\): keep the part for \(x \ge 0\) and mirror it in the y-axisModulus applied outside or inside the function.
\(y = k\,f(x)\) stretches vertically by factor \(k\); \(y = f(kx)\) compresses horizontally by factor \(k\)Scaling. For \(0 \lt k \lt 1\) the stretch and compression reverse.

Max, min, floor and ceiling

FormulaWhen to use
\(\max(a, b) = \dfrac{a + b + |a - b|}{2}\), \(\min(a, b) = \dfrac{a + b - |a - b|}{2}\)Rewriting max or min as an expression with modulus.
\(\max(a, b) + \min(a, b) = a + b\) and \(\max(a, b) \times \min(a, b) = ab\)Simplifying expressions that use both max and min of the same pair.
\(\lfloor x \rfloor \le x \lt \lfloor x \rfloor + 1\), and \(\{x\} = x - \lfloor x \rfloor\) with \(0 \le \{x\} \lt 1\)Turning a floor equation into an inequality on \(x\).
\(\lfloor x + n \rfloor = \lfloor x \rfloor + n\) for any integer \(n\); \(\lceil x \rceil = -\lfloor -x \rfloor\)Pulling integers out of a floor; converting ceiling to floor.
\(\lfloor x \rfloor + \lfloor -x \rfloor = 0\) if \(x\) is an integer, and \(-1\) otherwiseSums that pair \(x\) with \(-x\).

Shortcuts and traps

  • Floor of negatives: \(\lfloor -2.3 \rfloor = -3\), not \(-2\). The floor always moves left on the number line.
  • Special values first: in a functional equation try \(x = 0\), \(y = 0\), \(x = 1\) and \(y = -x\). For example, \(f(x + y) = f(x) + f(y)\) for all real \(x, y\) forces \(f(0) = 0\).
  • Composition order: \(f(g(x))\) applies \(g\) first. Check the domain of the inner function before simplifying.
  • Graphs of max and min: \(\max(f, g)\) is the upper envelope of the two graphs and \(\min(f, g)\) the lower one. The minimum of \(\max(f, g)\) is often where the two graphs cross, so check that point and the corners.

Inequalities & Modulus

Open on its own

Inequalities and modulus appear in CAT as range-of-values questions, minimum or maximum of an expression, and counting integer solutions. Focus on sign rules, AM-GM and the geometric meaning of \(|x - a|\) as a distance.

Rules for inequalities

FormulaWhen to use
If \(a \gt b\), then \(a + c \gt b + c\) for any real \(c\)Adding or subtracting the same quantity on both sides.
If \(a \gt b\): \(ac \gt bc\) when \(c \gt 0\), and \(ac \lt bc\) when \(c \lt 0\)Multiplying or dividing by a number. A negative number flips the sign.
If \(a \gt b\) and \(a, b\) have the same sign, then \(\dfrac{1}{a} \lt \dfrac{1}{b}\)Taking reciprocals. If \(a \gt 0 \gt b\), then \(\dfrac{1}{a} \gt \dfrac{1}{b}\) instead.
If \(a \gt b \ge 0\), then \(a^n \gt b^n\) and \(\sqrt{a} \gt \sqrt{b}\) for any positive integer \(n\)Squaring or taking roots. Safe only when both sides are non-negative.
For \(a \gt 0\): \(x^2 \lt a^2 \iff -a \lt x \lt a\), and \(x^2 \gt a^2 \iff x \lt -a\) or \(x \gt a\)Solving quadratic inequalities of the simplest form.

AM-GM

FormulaWhen to use
For \(a, b \ge 0\): \(\dfrac{a + b}{2} \ge \sqrt{ab}\), with equality only when \(a = b\)Bounding a sum by a product or a product by a sum.
For positive \(a_1, \dots, a_n\): AM \(\ge\) GM \(\ge\) HM, with equality only when all are equalComparing the three means of the same set of numbers.
Fixed sum: the product is largest when the numbers are equal. Fixed product: the sum is smallest when they are equalOptimisation with a constraint, for positive numbers.
\(x + \dfrac{1}{x} \ge 2\) for \(x \gt 0\); \(x + \dfrac{1}{x} \le -2\) for \(x \lt 0\)Range of expressions like \(x + \frac{1}{x}\) or \(\frac{x^2 + 1}{x}\).
For \(a, b \gt 0\) with \(a + b\) fixed and \(m, n \gt 0\), \(a^m b^n\) is largest when \(\dfrac{a}{m} = \dfrac{b}{n}\)Maximising a product with unequal powers, such as \(a^2 b\).

Modulus

FormulaWhen to use
\(|x| = x\) if \(x \ge 0\), \(|x| = -x\) if \(x \lt 0\); also \(|x| = \sqrt{x^2}\)Removing the modulus by splitting into cases.
\(|xy| = |x|\,|y|\) and \(\left|\dfrac{x}{y}\right| = \dfrac{|x|}{|y|}\) for \(y \neq 0\)Splitting a modulus over products and quotients. It does not split over sums.
For \(a \gt 0\): \(|x - c| \lt a \iff c - a \lt x \lt c + a\); \(|x - c| \gt a \iff x \lt c - a\) or \(x \gt c + a\)Solving modulus inequalities: distance of \(x\) from \(c\).
Triangle inequality: \(|a + b| \le |a| + |b|\), equality when \(ab \ge 0\); also \(|a - b| \ge \big||a| - |b|\big|\)Bounding a modulus of a sum or difference.
Minimum of \(|x - a| + |x - b|\) is \(|a - b|\), reached for every \(x\) between \(a\) and \(b\)Sum of two distances on the number line.
\(\sum |x - a_i|\) over \(n\) points is smallest at the median of the \(a_i\) (anywhere between the two middle points if \(n\) is even)Minimising a sum of three or more modulus terms.

Wavy-curve method

FormulaWhen to use
For \((x - a_1)(x - a_2)\cdots(x - a_n)\) with distinct roots and positive leading coefficient: the sign is \(+\) right of the largest root and alternates across each rootSolving polynomial inequalities such as \((x-1)(x+2)(x-4) \lt 0\).
A factor \((x - a)^k\): the sign changes across \(a\) if \(k\) is odd, and does not change if \(k\) is evenRepeated roots.
\(\dfrac{P(x)}{Q(x)} \ge 0\) has the same sign pattern as \(P(x)\,Q(x)\), but roots of \(Q\) are always excludedRational inequalities.

Shortcuts and traps

  • No blind cross-multiplying: in \(\dfrac{x}{x - 1} \gt 2\) the sign of \(x - 1\) is unknown. Move everything to one side, \(\dfrac{2 - x}{x - 1} \gt 0\), and use the wavy curve.
  • AM-GM needs positives: the bound holds only for non-negative terms, and it gives the true minimum or maximum only if equality can actually be reached.
  • Squaring both sides: allowed only when both sides are known to be non-negative. Otherwise it can add false solutions.
  • Endpoints: decide separately whether each critical point is included (\(\le\), \(\ge\)) or excluded (\(\lt\), \(\gt\), or a zero of a denominator).
  • Distance picture: read \(|x - a|\) as the distance from \(x\) to \(a\). Most modulus minimum questions are then solved by inspection.

Number System & Number Theory

Open on its own

Number System is tested in almost every CAT paper through factors, remainders, last digits and factorials. Learn prime factorisation well: nearly every formula below starts from \(N = p^a q^b r^c\), where \(p, q, r\) are distinct primes.

Divisibility and factors

FormulaWhen to use
Divisible by 2, 4, 8 if the last 1, 2, 3 digits are; by 3 or 9 if the digit sum is; by 11 if (sum of odd-place digits) \(-\) (sum of even-place digits) is a multiple of 11Quick divisibility checks.
Since \(1001 = 7 \times 11 \times 13\): split the number into groups of three digits from the right; it is divisible by 7, 11 or 13 if the alternating sum of the groups isTesting 7 and 13 on large numbers.
\(N\) is divisible by \(mn\) if it is divisible by both \(m\) and \(n\), provided \(\gcd(m, n) = 1\)Divisibility by 12, 15, 72 and so on: test \(4 \times 3\), \(3 \times 5\), \(8 \times 9\).
Number of factors of \(N\): \((a + 1)(b + 1)(c + 1)\)Counting all positive divisors.
Sum of factors: \(\dfrac{p^{a+1} - 1}{p - 1} \times \dfrac{q^{b+1} - 1}{q - 1} \times \dfrac{r^{c+1} - 1}{r - 1}\)Adding all positive divisors.
Product of factors: \(N^{d/2}\), where \(d\) is the number of factorsProduct of all positive divisors.
If \(N = 2^a \times M\) with \(M\) odd: odd factors \(=\) number of factors of \(M\); even factors \(=\) total \(-\) oddCounting odd or even divisors.
Ways to write \(N\) as a product of two factors (order ignored): \(\dfrac{d}{2}\) if \(N\) is not a perfect square, \(\dfrac{d + 1}{2}\) if it is. As a product of two coprime factors: \(2^{k-1}\), where \(k\) is the number of distinct primes"In how many ways can N be written as a product" questions.

HCF and LCM

FormulaWhen to use
\(\text{HCF}(a, b) \times \text{LCM}(a, b) = a \times b\)Finding one of the four values from the other three. Two numbers only.
HCF of fractions \(= \dfrac{\text{HCF of numerators}}{\text{LCM of denominators}}\); LCM of fractions \(= \dfrac{\text{LCM of numerators}}{\text{HCF of denominators}}\)Fractions in lowest terms.
Largest number that divides \(a, b, c\) leaving the same remainder: \(\text{HCF}(|a - b|, |b - c|)\)"Same remainder" questions.
Numbers leaving remainder \(r\) when divided by each of \(m, n, k\) (with \(r\) less than each divisor): \(t \times \text{LCM}(m, n, k) + r\), for integers \(t \ge 0\)"Same remainder each time" from the dividend side. The smallest such number above \(r\) is \(\text{LCM} + r\).

Remainders

FormulaWhen to use
Remainder of a sum or product \(=\) remainder of the sum or product of the individual remaindersBreaking a large expression into small pieces. Negative remainders such as \(-1\) are allowed.
Fermat: if \(p\) is prime and \(p\) does not divide \(a\), then \(a^{p-1}\) leaves remainder 1 when divided by \(p\)Large powers divided by a prime.
Euler: if \(\gcd(a, n) = 1\), then \(a^{\phi(n)}\) leaves remainder 1 when divided by \(n\), where \(\phi(n) = n\left(1 - \frac{1}{p}\right)\left(1 - \frac{1}{q}\right)\cdots\) over the distinct primes of \(n\)Large powers divided by a composite number.
Wilson: if \(p\) is prime, \((p - 1)!\) leaves remainder \(p - 1\) when divided by \(p\); also \((p - 2)!\) leaves remainder 1Factorials divided by a prime.
\(a^n - b^n\) is divisible by \(a - b\) for every positive integer \(n\), and by \(a + b\) when \(n\) is even; \(a^n + b^n\) is divisible by \(a + b\) when \(n\) is oddSpotting factors of expressions like \(17^{20} - 1\).

Factorials

FormulaWhen to use
Highest power of a prime \(p\) in \(n!\): \(\left\lfloor \dfrac{n}{p} \right\rfloor + \left\lfloor \dfrac{n}{p^2} \right\rfloor + \left\lfloor \dfrac{n}{p^3} \right\rfloor + \cdots\)Largest \(k\) such that \(p^k\) divides \(n!\).
Trailing zeros of \(n!\): \(\left\lfloor \dfrac{n}{5} \right\rfloor + \left\lfloor \dfrac{n}{25} \right\rfloor + \left\lfloor \dfrac{n}{125} \right\rfloor + \cdots\)Number of zeros at the end of a factorial (fives are scarcer than twos).
For a composite base such as \(12 = 2^2 \times 3\): find the power of each prime in \(n!\), divide by its exponent in the base, take the floor, and use the smallest resultHighest power of a composite number in \(n!\).

Last-digit cycles

Last digit of baseCycle of last digitsLast digit of baseCycle of last digits
22, 4, 8, 677, 9, 3, 1
33, 9, 7, 188, 4, 2, 6
44, 699, 1
0, 1, 5, 6Always the same digit

Shortcuts and traps

  • Using the cycle: for an exponent \(n \ge 1\), find the remainder of \(n\) divided by 4; a remainder of 0 means take the 4th term. So \(7^{100}\) ends in 1.
  • Use \(-1\): \(2^{100}\) divided by 3: since \(2\) leaves \(-1\), the remainder is \((-1)^{100} = 1\).
  • Coprime condition: Fermat and Euler fail if the base shares a factor with the divisor. Split the divisor into coprime parts first.
  • HCF \(\times\) LCM: the product rule works for two numbers only, never for three.
  • 1 is special: 1 is neither prime nor composite, and it is counted among the factors of every number.

Geometry Basics

Open on its own

Basic angle facts rarely make a full CAT question on their own, but they are the first step in most Geometry questions. Focus on parallel-line angle pairs, polygon angle sums and the conditions for congruence and similarity.

Lines and angles

FormulaWhen to use
Angles on a straight line add to \(180^\circ\); angles around a point add to \(360^\circ\)Finding a missing angle at a point or on a line.
Vertically opposite angles are equalTwo straight lines crossing each other.
Complementary angles add to \(90^\circ\); supplementary angles add to \(180^\circ\)Word statements such as "an angle is twice its complement".
Parallel lines cut by a transversal: corresponding angles are equal, alternate interior angles are equal, alternate exterior angles are equalCarrying an angle from one parallel line to the other.
Parallel lines cut by a transversal: co-interior (same-side interior) angles add to \(180^\circ\)Angles between the parallel lines on the same side of the transversal.
If any one of the transversal angle relations above (corresponding, alternate or co-interior) holds for two lines and a transversal, the two lines are parallelProving or using that two lines are parallel.

Angle sums in triangles and polygons

FormulaWhen to use
Angles of a triangle add to \(180^\circ\); an exterior angle equals the sum of the two interior opposite anglesAny triangle with one or two angles known.
Sum of interior angles of an \(n\)-sided polygon: \((n - 2) \times 180^\circ\)Quadrilaterals (\(360^\circ\)), pentagons (\(540^\circ\)) and so on.
Sum of exterior angles of a convex polygon, one at each vertex: \(360^\circ\)Holds for every convex polygon, whatever the number of sides.
Regular \(n\)-gon: each exterior angle \(= \dfrac{360^\circ}{n}\), each interior angle \(= \dfrac{(n - 2) \times 180^\circ}{n}\)Finding \(n\) from one angle, or one angle from \(n\).
Number of diagonals of an \(n\)-sided polygon: \(\dfrac{n(n - 3)}{2}\)Counting diagonals or finding \(n\) from the count.

Congruence and similarity

FormulaWhen to use
Congruence tests: SSS, SAS (included angle), ASA, AAS, RHS (right angle, hypotenuse, one side)Showing two triangles are identical in shape and size.
Similarity tests: AA, SSS (all three sides in the same ratio), SAS (two sides in the same ratio and the included angle equal)Showing two triangles have the same shape.
Similar figures with side ratio \(k\): lengths in ratio \(k\), areas in ratio \(k^2\), volumes in ratio \(k^3\)Comparing areas or volumes of similar shapes.
Basic proportionality: in \(\triangle ABC\), if \(DE \parallel BC\) with \(D\) on \(AB\) and \(E\) on \(AC\), then \(\dfrac{AD}{DB} = \dfrac{AE}{EC}\)A line parallel to one side of a triangle.
Three or more parallel lines cut any two transversals in the same ratioProportional segments between parallel lines.

Regular polygons

SidesInterior angleSidesInterior angle
3\(60^\circ\)8\(135^\circ\)
4\(90^\circ\)9\(140^\circ\)
5\(108^\circ\)10\(144^\circ\)
6\(120^\circ\)12\(150^\circ\)

Shortcuts and traps

  • SSA and AAA are not congruence tests: two sides and a non-included angle can give two different triangles, and equal angles only give similarity.
  • Work from the exterior angle: for regular polygons, \(\frac{360^\circ}{\text{exterior angle}}\) must be a whole number, which quickly rules out impossible angles.
  • Similar areas: doubling every length multiplies the area by 4, not 2.
  • Corresponding sides: match sides opposite equal angles before writing a similarity ratio. The order of letters in \(\triangle ABC \sim \triangle PQR\) gives the matching.

Triangles

Open on its own

Triangles are the most tested part of CAT Geometry, either directly or hidden inside circle and coordinate questions. Focus on area formulas, Pythagorean triplets, similarity and the special right triangles; below, \(a, b, c\) are the sides opposite angles \(A, B, C\), and \(s = \dfrac{a + b + c}{2}\).

Area and radii

FormulaWhen to use
Area \(= \dfrac{1}{2} \times \text{base} \times \text{height} = \dfrac{1}{2}\,ab \sin C\)Height known, or two sides and the included angle known.
Heron: Area \(= \sqrt{s(s - a)(s - b)(s - c)}\)All three sides known.
Area \(= r \times s\) and Area \(= \dfrac{abc}{4R}\), where \(r\) is the inradius and \(R\) the circumradiusFinding \(r\) or \(R\) once the area is known.
Equilateral triangle of side \(a\): area \(= \dfrac{\sqrt{3}}{4}a^2\), height \(= \dfrac{\sqrt{3}}{2}a\), \(R = \dfrac{a}{\sqrt{3}}\), \(r = \dfrac{a}{2\sqrt{3}}\)Any equilateral triangle. Note \(R = 2r\).
Right triangle with legs \(a, b\) and hypotenuse \(c\): \(R = \dfrac{c}{2}\), \(r = \dfrac{a + b - c}{2}\)Circles inside or around a right triangle.
Sine rule: \(\dfrac{a}{\sin A} = \dfrac{b}{\sin B} = \dfrac{c}{\sin C} = 2R\)Linking sides, opposite angles and the circumradius.

Sides and lengths

FormulaWhen to use
Triangle inequality: \(|b - c| \lt a \lt b + c\) for every sideChecking whether three lengths form a triangle, or the range of a third side.
With \(c\) the longest side: \(c^2 = a^2 + b^2\) means right-angled, \(c^2 \gt a^2 + b^2\) obtuse, \(c^2 \lt a^2 + b^2\) acuteClassifying a triangle from its sides.
Triplets: \((3, 4, 5)\), \((5, 12, 13)\), \((8, 15, 17)\), \((7, 24, 25)\), \((20, 21, 29)\), \((9, 40, 41)\) and their multiples. In general \((m^2 - n^2,\ 2mn,\ m^2 + n^2)\) for integers \(m \gt n \gt 0\)Spotting integer right triangles without calculation.
Angle bisector from \(A\) meeting \(BC\) at \(D\): \(\dfrac{BD}{DC} = \dfrac{AB}{AC} = \dfrac{c}{b}\); length \(AD^2 = bc\left[1 - \dfrac{a^2}{(b + c)^2}\right]\)A line that bisects an angle of a triangle.
Apollonius: if \(D\) is the midpoint of \(BC\), \(AB^2 + AC^2 = 2(AD^2 + BD^2)\), so median \(m_a = \dfrac{1}{2}\sqrt{2b^2 + 2c^2 - a^2}\)Finding the length of a median.
Right angle at \(A\), altitude \(AD\) to the hypotenuse: \(AD = \dfrac{AB \times AC}{BC}\) and \(AD^2 = BD \times DC\)Altitude drawn to the hypotenuse of a right triangle.

Special triangles and similarity

FormulaWhen to use
\(30^\circ\)-\(60^\circ\)-\(90^\circ\): sides opposite these angles are in ratio \(1 : \sqrt{3} : 2\)Half of an equilateral triangle, or any triangle with a \(30^\circ\) or \(60^\circ\) angle and a right angle.
\(45^\circ\)-\(45^\circ\)-\(90^\circ\): sides in ratio \(1 : 1 : \sqrt{2}\)Half of a square cut along a diagonal.
Similar triangles: ratio of areas \(=\) square of the ratio of corresponding sides (also of corresponding heights, medians or bisectors)Area of a smaller triangle cut off by a parallel line.

Centres of a triangle

FormulaWhen to use
Centroid \(G\) (meet of medians) divides each median in ratio \(2 : 1\) from the vertex; the three medians split the triangle into 6 triangles of equal areaLengths along medians and areas of pieces.
Incentre \(I\) (meet of angle bisectors) is equidistant from all three sides; \(\angle BIC = 90^\circ + \dfrac{A}{2}\)Angles at the incentre and the inscribed circle.
Circumcentre \(O\) (meet of perpendicular bisectors) is equidistant from all three vertices; \(\angle BOC = 2A\) when \(A\) is acute. In a right triangle it is the midpoint of the hypotenuseAngles at the circumcentre and the circumscribed circle.

Shortcuts and traps

  • Strict inequality: the triangle inequality is strict. Sides 2, 3, 5 do not form a triangle.
  • Longest side only: compare \(c^2\) with \(a^2 + b^2\) only for the longest side \(c\); for the other sides the test tells you nothing.
  • Equilateral coincidence: in an equilateral triangle the centroid, incentre, circumcentre and orthocentre are the same point, and it lies \(\frac{1}{3}\) of the height above the base.
  • Scale triplets: a right triangle with sides 15 and 20 has hypotenuse 25, since it is \(5 \times (3, 4, 5)\).

Circles

Open on its own

Circle questions in CAT usually combine an angle property with a length result such as a chord distance, a tangent length or the power of a point. Know which angle facts apply to chords and which to tangents, and draw the radius to every point of contact.

Measurement

FormulaWhen to use
Circumference \(= 2\pi r\); area \(= \pi r^2\)Perimeter and area of a full circle of radius \(r\).
Arc length \(= \dfrac{\theta}{360^\circ} \times 2\pi r\); sector area \(= \dfrac{\theta}{360^\circ} \times \pi r^2 = \dfrac{1}{2} \times \text{arc} \times r\)Part of a circle cut off by a central angle \(\theta\).
Segment area \(=\) sector area \(-\) triangle area \(= \dfrac{\theta}{360^\circ} \times \pi r^2 - \dfrac{1}{2} r^2 \sin\theta\)Region between a chord and its minor arc.
Ring area \(= \pi(R^2 - r^2)\) for concentric circles of radii \(R \gt r\)Paths, tracks and washers.

Chords and angles

FormulaWhen to use
The perpendicular from the centre bisects a chord; a chord at distance \(d\) from the centre has length \(2\sqrt{r^2 - d^2}\)Linking chord length, radius and distance from the centre.
Equal chords are equidistant from the centre and subtend equal angles at the centreComparing chords in the same circle.
Angle at the centre \(= 2 \times\) angle at the circumference on the same arcMoving between central and inscribed angles.
Angles in the same segment are equal; the angle in a semicircle is \(90^\circ\)Several inscribed angles on one chord, or a diameter as a side.
Cyclic quadrilateral: opposite angles add to \(180^\circ\); an exterior angle equals the interior opposite angleFour vertices on one circle.
Cyclic quadrilateral with sides \(a, b, c, d\) and \(s = \dfrac{a + b + c + d}{2}\): area \(= \sqrt{(s - a)(s - b)(s - c)(s - d)}\)Area of a cyclic quadrilateral from its sides only.

Tangents and power of a point

FormulaWhen to use
A tangent is perpendicular to the radius at the point of contactCreating a right angle to apply Pythagoras.
The two tangents from an external point \(P\) are equal, each of length \(\sqrt{d^2 - r^2}\), where \(d\) is the distance from \(P\) to the centreTangent lengths from one point.
Alternate segment: the angle between a tangent and a chord equals the inscribed angle on that chord in the opposite segmentAngles where a tangent meets a chord.
Quadrilateral \(ABCD\) with a circle touching all four sides: \(AB + CD = BC + DA\)Quadrilaterals drawn around a circle.
Chords \(AB\) and \(CD\) meeting at \(P\) inside the circle: \(PA \times PB = PC \times PD\)Two chords that cross.
Secants \(PAB\) and \(PCD\) from an external point \(P\): \(PA \times PB = PC \times PD\); with a tangent \(PT\): \(PT^2 = PA \times PB\)Lines from an outside point that cut or touch the circle.

Two circles

FormulaWhen to use
Direct common tangent length \(= \sqrt{d^2 - (r_1 - r_2)^2}\), where \(d\) is the distance between centresTangent that does not pass between the circles.
Transverse common tangent length \(= \sqrt{d^2 - (r_1 + r_2)^2}\), for \(d \gt r_1 + r_2\)Tangent that crosses the line joining the centres.

Number of common tangents (\(r_1 \ge r_2\))

PositionConditionCommon tangents
Separate\(d \gt r_1 + r_2\)4
Touching externally\(d = r_1 + r_2\)3
Intersecting\(r_1 - r_2 \lt d \lt r_1 + r_2\)2
Touching internally\(d = r_1 - r_2\), \(d \gt 0\)1
One inside the other\(d \lt r_1 - r_2\)0

Shortcuts and traps

  • Same arc, same side: the centre-angle rule needs both angles to stand on the same arc. An inscribed angle on the major arc is \(180^\circ\) minus the one on the minor arc.
  • Whole secant, not the inside part: in \(PA \times PB\), both lengths are measured from \(P\). Using the chord \(AB\) instead of \(PB\) is the most common error.
  • Draw the radius: joining the centre to a point of contact or to the ends of a chord almost always creates the right triangle you need.
  • Brahmagupta needs a cyclic quadrilateral: for a general quadrilateral the side lengths alone do not fix the area.

Quadrilaterals & Polygons

Open on its own

Quadrilaterals and polygons appear in CAT Geometry as area, diagonal and angle questions, often mixed with triangles and circles. Know the area formula for each shape and the angle and diagonal counts for regular polygons.

Quadrilaterals

FormulaWhen to use
Square of side \(a\): area \(= a^2 = \dfrac{d^2}{2}\), diagonal \(d = a\sqrt{2}\)Square questions where either the side or the diagonal is given.
Rectangle: area \(= lb\), perimeter \(= 2(l + b)\), diagonal \(= \sqrt{l^2 + b^2}\)Any rectangle; the diagonal comes from Pythagoras.
Parallelogram: area \(= \text{base} \times \text{height} = ab\sin\theta\)Adjacent sides \(a, b\) with included angle \(\theta\), or a base with its perpendicular height.
Parallelogram diagonals: \(d_1^2 + d_2^2 = 2(a^2 + b^2)\)Finding a diagonal or side when the other three lengths are known. Holds for rectangles, rhombuses and squares too.
Rhombus: area \(= \dfrac{1}{2} d_1 d_2\), side \(a\) satisfies \(4a^2 = d_1^2 + d_2^2\)The diagonals of a rhombus bisect each other at right angles.
Trapezium: area \(= \dfrac{1}{2}(a + b) \times h\)\(a\) and \(b\) are the parallel sides and \(h\) is the distance between them.
Trapezium mid-segment \(= \dfrac{a + b}{2}\)The segment joining the midpoints of the non-parallel sides is parallel to \(a\) and \(b\).
Kite: area \(= \dfrac{1}{2} d_1 d_2\)The diagonals of a kite are perpendicular; only one of them bisects the other.
Any quadrilateral: area \(= \dfrac{1}{2} d_1 d_2 \sin\theta\)Diagonals \(d_1, d_2\) meeting at angle \(\theta\). Becomes \(\frac{1}{2}d_1 d_2\) when they are perpendicular.
Cyclic quadrilateral: opposite angles sum to \(180^\circ\); area \(= \sqrt{(s - a)(s - b)(s - c)(s - d)}\), \(s = \dfrac{a + b + c + d}{2}\)Only when all four vertices lie on a circle (Brahmagupta's formula).

Polygons

FormulaWhen to use
Sum of interior angles \(= (n - 2) \times 180^\circ\)Any convex polygon with \(n\) sides.
Sum of exterior angles \(= 360^\circ\)Any convex polygon, whatever the number of sides.
Regular polygon: each exterior angle \(= \dfrac{360^\circ}{n}\), each interior angle \(= \dfrac{(n - 2) \times 180^\circ}{n}\)Only for regular polygons. Use the exterior angle to find \(n\) quickly.
Number of diagonals \(= \dfrac{n(n - 3)}{2}\)Any polygon with \(n\) sides.
Regular polygon area \(= \dfrac{1}{2} \times \text{perimeter} \times \text{inradius}\)The inradius (apothem) is the distance from the centre to any side.
Regular hexagon of side \(a\): area \(= \dfrac{3\sqrt{3}}{2}a^2\), longest diagonal \(= 2a\), shorter diagonal \(= \sqrt{3}a\)The hexagon splits into six equilateral triangles of side \(a\).
Regular octagon of side \(a\): area \(= 2(1 + \sqrt{2})a^2\)Octagon cut from a square by removing four corner triangles.

Shortcuts and traps

  • Sides are not enough: a parallelogram or rhombus is not fixed by its sides alone; you also need an angle, a height or a diagonal.
  • Fixed perimeter: among all rectangles with the same perimeter, the square has the largest area.
  • Find n from angles: if each interior angle of a regular polygon is \(x^\circ\), then \(n = \dfrac{360}{180 - x}\).
  • Hexagon as triangles: break a regular hexagon into six equilateral triangles, or into a rectangle and two triangles, instead of memorising extra results.
  • Cyclic only: Brahmagupta's formula and the "opposite angles sum to \(180^\circ\)" rule do not apply to a general quadrilateral.

Mensuration

Open on its own

Mensuration questions in CAT ask for volumes and surface areas of standard solids, usually with a twist such as melting one solid into another or cutting a solid. Learn each formula with its condition and always keep units consistent.

Cube and cuboid

FormulaWhen to use
Cube of side \(a\): volume \(= a^3\), lateral surface \(= 4a^2\), total surface \(= 6a^2\)Any cube.
Cube space diagonal \(= a\sqrt{3}\); face diagonal \(= a\sqrt{2}\)Longest rod that fits in a cubical box, or distance between opposite corners.
Cuboid: volume \(= lbh\), total surface \(= 2(lb + bh + hl)\), lateral surface \(= 2h(l + b)\)Boxes, rooms and tanks. Lateral surface means the four walls only.
Cuboid space diagonal \(= \sqrt{l^2 + b^2 + h^2}\)Longest rod that fits in a room or box.
\((l + b + h)^2 = (\text{diagonal})^2 + \text{total surface area}\)When the sum of edges and the diagonal or surface area are given.

Cylinder, cone and frustum

FormulaWhen to use
Cylinder: volume \(= \pi r^2 h\), curved surface \(= 2\pi rh\), total surface \(= 2\pi r(r + h)\)Right circular cylinder of radius \(r\) and height \(h\).
Hollow cylinder: volume \(= \pi h(R^2 - r^2)\)Pipes and tubes with outer radius \(R\) and inner radius \(r\).
Cone slant height \(l = \sqrt{r^2 + h^2}\)Right circular cone; needed before any cone surface area.
Cone: volume \(= \dfrac{1}{3}\pi r^2 h\), curved surface \(= \pi rl\), total surface \(= \pi r(l + r)\)Right circular cone. Its volume is one third of the cylinder with the same base and height.
Frustum: volume \(= \dfrac{1}{3}\pi h(R^2 + r^2 + Rr)\)A cone with its top cut off parallel to the base; \(R\) and \(r\) are the end radii.
Frustum: slant \(l = \sqrt{h^2 + (R - r)^2}\), curved surface \(= \pi(R + r)l\), total surface \(= \pi(R + r)l + \pi R^2 + \pi r^2\)Buckets and lampshades. Drop the \(\pi r^2\) or \(\pi R^2\) term for an open end.
Cone from a sector: sector radius \(=\) slant height \(l\); arc length \(= 2\pi r\)Folding a sector of a circle into a cone.

Sphere and hemisphere

FormulaWhen to use
Sphere: volume \(= \dfrac{4}{3}\pi r^3\), surface area \(= 4\pi r^2\)Balls and spherical shells.
Hemisphere: volume \(= \dfrac{2}{3}\pi r^3\), curved surface \(= 2\pi r^2\), total surface \(= 3\pi r^2\)Solid hemisphere; total surface includes the flat circular face.
Spherical shell: volume \(= \dfrac{4}{3}\pi(R^3 - r^3)\)Hollow ball with outer radius \(R\) and inner radius \(r\).

Shortcuts and traps

  • Melting and recasting: volume is conserved, surface area is not. Number of small solids \(= \dfrac{\text{volume of big solid}}{\text{volume of one small solid}}\).
  • Scaling: if every length is multiplied by \(k\), surface areas multiply by \(k^2\) and volumes by \(k^3\).
  • Cutting adds surface: cutting a solid keeps the total volume but creates new faces, so total surface area goes up.
  • Curved versus total: read whether the question wants curved (lateral) or total surface area, and whether a container is open at the top.
  • Units: \(1 \text{ m}^3 = 1000\) litres and \(1 \text{ litre} = 1000 \text{ cm}^3\).

Trigonometry

Open on its own

Trigonometry in CAT is light: standard values, a few identities and heights-and-distances problems, plus the sine and cosine rules inside triangle questions. Memorise the values table and the two special right triangles.

Ratios and identities

FormulaWhen to use
\(\sin\theta = \dfrac{\text{opposite}}{\text{hypotenuse}}\), \(\cos\theta = \dfrac{\text{adjacent}}{\text{hypotenuse}}\), \(\tan\theta = \dfrac{\text{opposite}}{\text{adjacent}}\)Any acute angle \(\theta\) in a right-angled triangle.
\(\csc\theta = \dfrac{1}{\sin\theta}\), \(\sec\theta = \dfrac{1}{\cos\theta}\), \(\cot\theta = \dfrac{1}{\tan\theta}\), \(\tan\theta = \dfrac{\sin\theta}{\cos\theta}\)Converting between ratios. Each holds where the denominator is not zero.
\(\sin^2\theta + \cos^2\theta = 1\)Finding one ratio from another; true for every angle.
\(1 + \tan^2\theta = \sec^2\theta\), \(1 + \cot^2\theta = \csc^2\theta\)Simplifying expressions in tan/sec or cot/csc.
\(\sin(90^\circ - \theta) = \cos\theta\), \(\tan(90^\circ - \theta) = \cot\theta\), \(\sec(90^\circ - \theta) = \csc\theta\)Complementary angles, for example \(\tan 20^\circ \times \tan 70^\circ = 1\).
\(\sin 2\theta = 2\sin\theta\cos\theta\), \(\cos 2\theta = \cos^2\theta - \sin^2\theta\)Double-angle simplification.

Standard values

Ratio\(0^\circ\)\(30^\circ\)\(45^\circ\)\(60^\circ\)\(90^\circ\)
\(\sin\)0\(\frac{1}{2}\)\(\frac{1}{\sqrt{2}}\)\(\frac{\sqrt{3}}{2}\)1
\(\cos\)1\(\frac{\sqrt{3}}{2}\)\(\frac{1}{\sqrt{2}}\)\(\frac{1}{2}\)0
\(\tan\)0\(\frac{1}{\sqrt{3}}\)1\(\sqrt{3}\)not defined

Triangles, heights and distances

FormulaWhen to use
Sides of a \(30^\circ\)-\(60^\circ\)-\(90^\circ\) triangle are in the ratio \(1 : \sqrt{3} : 2\); of a \(45^\circ\)-\(45^\circ\)-\(90^\circ\) triangle, \(1 : 1 : \sqrt{2}\)Most heights-and-distances questions reduce to one of these.
Height \(= d \tan\theta\)Angle of elevation \(\theta\) to the top of an object from a point at horizontal distance \(d\).
Height \(= \dfrac{d}{\cot\alpha - \cot\beta}\)Elevations \(\alpha < \beta\) from two points on the same side, \(d\) apart, in line with the foot of the object.
Sine rule: \(\dfrac{a}{\sin A} = \dfrac{b}{\sin B} = \dfrac{c}{\sin C} = 2R\)Any triangle, when an angle and its opposite side are known. \(R\) is the circumradius.
Cosine rule: \(c^2 = a^2 + b^2 - 2ab\cos C\), so \(\cos C = \dfrac{a^2 + b^2 - c^2}{2ab}\)Any triangle, given two sides and the included angle, or all three sides.
Area \(= \dfrac{1}{2}ab\sin C\)Two sides and the angle between them.

Shortcuts and traps

  • Elevation equals depression: the angle of depression from A to B equals the angle of elevation from B to A, because they are alternate angles.
  • Not linear: \(\sin(A + B) \neq \sin A + \sin B\) and \(\sin 2\theta \neq 2\sin\theta\) in general.
  • Maximum value: \(a\sin\theta + b\cos\theta\) ranges from \(-\sqrt{a^2 + b^2}\) to \(\sqrt{a^2 + b^2}\).
  • Ambiguous case: given two sides and a non-included angle, the sine rule can give two different triangles; check whether the obtuse angle also fits.
  • Draw first: sketch the figure and mark the right angle before choosing a ratio; most errors come from using the wrong side as "opposite".

Coordinate Geometry

Open on its own

Coordinate Geometry questions in CAT test distances, areas, slopes and straight lines, and occasionally the equation of a circle. Plot the points roughly before calculating; a sketch often rules out options straight away.

Points

FormulaWhen to use
Distance \(= \sqrt{(x_2 - x_1)^2 + (y_2 - y_1)^2}\)Length between two points; also to test for isosceles or right triangles.
Internal division in ratio \(m : n\): \(\left(\dfrac{mx_2 + nx_1}{m + n}, \dfrac{my_2 + ny_1}{m + n}\right)\)Point on segment from \((x_1, y_1)\) to \((x_2, y_2)\) with \(AP : PB = m : n\).
External division in ratio \(m : n\): \(\left(\dfrac{mx_2 - nx_1}{m - n}, \dfrac{my_2 - ny_1}{m - n}\right)\)Point outside the segment with \(AP : PB = m : n\), where \(m \neq n\).
Midpoint \(= \left(\dfrac{x_1 + x_2}{2}, \dfrac{y_1 + y_2}{2}\right)\)The \(1 : 1\) case; also, diagonals of a parallelogram share a midpoint.
Centroid \(= \left(\dfrac{x_1 + x_2 + x_3}{3}, \dfrac{y_1 + y_2 + y_3}{3}\right)\)Point where the medians of a triangle meet.
Area of triangle \(= \dfrac{1}{2}\left|x_1(y_2 - y_3) + x_2(y_3 - y_1) + x_3(y_1 - y_2)\right|\)Three vertices known. Area 0 means the points are collinear.

Straight lines

FormulaWhen to use
Slope \(m = \dfrac{y_2 - y_1}{x_2 - x_1} = \tan\theta\)Line through two points with \(x_1 \neq x_2\); \(\theta\) is the angle with the positive x-axis.
Point-slope: \(y - y_1 = m(x - x_1)\); slope-intercept: \(y = mx + c\)Writing a line from a point and a slope, or from the slope and y-intercept \(c\).
Intercept form: \(\dfrac{x}{a} + \dfrac{y}{b} = 1\); area with the axes \(= \dfrac{1}{2}|ab|\)Line cutting the x-axis at \(a\) and the y-axis at \(b\), both non-zero.
General form \(ax + by + c = 0\): slope \(= -\dfrac{a}{b}\)Reading the slope directly, when \(b \neq 0\).
Parallel: \(m_1 = m_2\); perpendicular: \(m_1 m_2 = -1\)Neither line vertical. A vertical line is perpendicular to any horizontal line.
Angle between lines: \(\tan\theta = \left|\dfrac{m_1 - m_2}{1 + m_1 m_2}\right|\)Acute angle between two lines, when \(m_1 m_2 \neq -1\).
Distance from \((x_1, y_1)\) to \(ax + by + c = 0\): \(\dfrac{|ax_1 + by_1 + c|}{\sqrt{a^2 + b^2}}\)Perpendicular distance from a point to a line.
Distance between \(ax + by + c_1 = 0\) and \(ax + by + c_2 = 0\): \(\dfrac{|c_1 - c_2|}{\sqrt{a^2 + b^2}}\)Parallel lines written with the same \(a\) and \(b\).

Circle

FormulaWhen to use
\((x - h)^2 + (y - k)^2 = r^2\)Circle with centre \((h, k)\) and radius \(r\).
\(x^2 + y^2 + 2gx + 2fy + c = 0\): centre \((-g, -f)\), radius \(= \sqrt{g^2 + f^2 - c}\)General form; it is a real circle only when \(g^2 + f^2 - c > 0\).

Shortcuts and traps

  • Ratio order: in the section formula, \(m\) multiplies the coordinates of the second point; swapping them gives the wrong point.
  • Match coefficients first: before using the parallel-lines distance formula, rewrite both lines so \(a\) and \(b\) are identical (for example, \(2x + 4y = 3\) becomes \(x + 2y = 1.5\)).
  • Vertical lines: a vertical line has no slope, so \(m_1 m_2 = -1\) cannot be used; check vertical and horizontal cases separately.
  • Absolute value: area and distance come out with a sign from the formula; always take the absolute value.
  • Collinearity: three points are collinear if the triangle area is zero, or if two pairs of them have the same slope.

Permutation & Combination

Open on its own

Permutation and Combination is a regular source of CAT questions, often as TITA. Decide first whether order matters and whether the objects and the boxes are identical or distinct; the right formula follows from those two answers.

Core counting

FormulaWhen to use
Multiplication rule: \(m \times n\) ways; addition rule: \(m + n\) waysMultiply for steps done one after another ("and"); add for separate cases ("or").
\(n! = n \times (n - 1) \times \dots \times 1\), with \(0! = 1\)Arranging all \(n\) distinct objects in a row.
\({}^nP_r = \dfrac{n!}{(n - r)!}\)Arranging \(r\) out of \(n\) distinct objects, order matters, \(0 \le r \le n\).
\({}^nC_r = \binom{n}{r} = \dfrac{n!}{r!\,(n - r)!}\)Selecting \(r\) out of \(n\) distinct objects, order does not matter, \(0 \le r \le n\).
\(\binom{n}{r} = \binom{n}{n - r}\); \(\binom{n}{r} + \binom{n}{r - 1} = \binom{n + 1}{r}\)Simplifying selections, for \(1 \le r \le n\).
\(\dfrac{n!}{p!\,q!\,r!\cdots}\)Arranging \(n\) objects in a row where \(p\) are alike of one kind, \(q\) of another, and so on (letters of a word).
\(n^r\)Arrangements of length \(r\) from \(n\) distinct objects when repetition is allowed.

Arrangements and selections

FormulaWhen to use
Circular arrangements \(= (n - 1)!\)\(n\) distinct people around a round table, where only relative positions matter.
Necklace or garland \(= \dfrac{(n - 1)!}{2}\)\(n \ge 3\) distinct beads, when clockwise and anticlockwise orders count as the same.
\(2^n - 1\)Selecting at least one object from \(n\) distinct objects.
\((p + 1)(q + 1)(r + 1) - 1\)Selecting at least one object from \(p\) identical, \(q\) identical and \(r\) identical objects of different kinds.
Handshakes or lines \(= \binom{n}{2}\); triangles \(= \binom{n}{3}\); polygon diagonals \(= \binom{n}{2} - n\)\(n\) people or \(n\) points with no three collinear.
Grid paths \(= \binom{m + n}{m}\)Shortest paths on a grid moving only right or up, \(m\) steps right and \(n\) steps up.
Derangements \(D_n = n!\left(1 - \dfrac{1}{1!} + \dfrac{1}{2!} - \dots + \dfrac{(-1)^n}{n!}\right)\); \(D_2 = 1, D_3 = 2, D_4 = 9, D_5 = 44\)Ways to put \(n\) letters in \(n\) envelopes so that none is in its correct envelope.

Distribution

FormulaWhen to use
\(\binom{n + r - 1}{r - 1}\)\(n\) identical objects into \(r\) distinct boxes, empty boxes allowed. Same as non-negative integer solutions of \(x_1 + \dots + x_r = n\).
\(\binom{n - 1}{r - 1}\)\(n\) identical objects into \(r\) distinct boxes, each box non-empty (\(n \ge r\)). Same as positive integer solutions of \(x_1 + \dots + x_r = n\).
\(r^n\)\(n\) distinct objects into \(r\) distinct boxes, any box may be empty (each object picks a box).
\(\dfrac{(mn)!}{(n!)^m \, m!}\)Splitting \(mn\) distinct objects into \(m\) unlabelled groups of \(n\) each. Drop the \(m!\) if the groups are labelled.

Shortcuts and traps

  • Order or not: "arrange", "rank", "form numbers" means permutations; "choose", "select", "form a team" means combinations.
  • Together: tie the objects that must be together into one unit, arrange the units, then multiply by the arrangements inside the unit.
  • Never together: arrange the other objects first, then place the restricted ones in the gaps between them (gap method).
  • At least one: count the total and subtract the cases with none, instead of adding many cases.
  • Lower limits: for \(x_i \ge k\), substitute \(y_i = x_i - k\) to get a non-negative problem, then use \(\binom{n + r - 1}{r - 1}\) on the new total.

Probability

Open on its own

Probability questions in CAT build on counting, so most of the work is getting the favourable and total cases right. Focus on the complement rule, independent events and conditional probability.

Basic rules

FormulaWhen to use
\(P(E) = \dfrac{\text{favourable outcomes}}{\text{total outcomes}}\), with \(0 \le P(E) \le 1\)Only when all outcomes are equally likely.
\(P(\text{not } E) = 1 - P(E)\)When the opposite event is easier to count.
\(P(A \cup B) = P(A) + P(B) - P(A \cap B)\)Probability of A or B (or both).
\(P(A \cup B) = P(A) + P(B)\)Only when A and B are mutually exclusive (cannot happen together).
\(P(A \cup B \cup C) = P(A) + P(B) + P(C) - P(A \cap B) - P(B \cap C) - P(A \cap C) + P(A \cap B \cap C)\)Three events that may overlap.
Odds in favour \(a : b\) means \(P = \dfrac{a}{a + b}\); odds against \(a : b\) means \(P = \dfrac{b}{a + b}\)Converting between odds and probability.

Dependent and independent events

FormulaWhen to use
\(P(A \cap B) = P(A) \times P(B)\)Only when A and B are independent (one does not affect the other).
\(P(A \mid B) = \dfrac{P(A \cap B)}{P(B)}\)Probability of A given that B has happened, with \(P(B) > 0\).
\(P(A \cap B) = P(A) \times P(B \mid A)\)Events in sequence, such as drawing without replacement.
\(P(A \mid B) = \dfrac{P(B \mid A)\,P(A)}{P(B)}\), where \(P(B) = P(B \mid A)P(A) + P(B \mid A')P(A')\)Reversing a condition (Bayes' rule), for example "given the ball is red, which bag was it from".
\(P(\text{at least one}) = 1 - (1 - p_1)(1 - p_2)\cdots(1 - p_n)\)Independent events; becomes \(1 - (1 - p)^n\) when all probabilities equal \(p\).
\(P(\text{exactly } r \text{ successes}) = \binom{n}{r} p^r (1 - p)^{n - r}\)\(n\) independent repeated trials, each with the same success probability \(p\).
Expected value \(= \sum x_i \, p_i\)Average outcome of a game or bet with values \(x_i\) and probabilities \(p_i\).

Standard facts

ExperimentFact
\(n\) coins tossed\(2^n\) outcomes; \(P(\text{exactly } r \text{ heads}) = \dfrac{\binom{n}{r}}{2^n}\)
\(n\) dice rolled\(6^n\) outcomes
Two dice, sum \(s\) (2 to 12)\(6 - |s - 7|\) ways out of 36; sum 7 is the most likely, with 6 ways
Deck of 52 cards4 suits of 13; 26 red and 26 black; 12 face cards (J, Q, K); 4 aces

Shortcuts and traps

  • Exclusive is not independent: two events with non-zero probability that are mutually exclusive cannot be independent, because one happening rules out the other.
  • Without replacement: after each draw the total and the favourable count both change; multiply the step-by-step probabilities.
  • At least one: use \(1 - P(\text{none})\); adding cases one by one is slow and easy to get wrong.
  • Equally likely check: the sums of two dice are not equally likely, so "11 possible sums" does not give a probability of \(\frac{1}{11}\) each.
  • Given means shrink: in a conditional question, the "given" event becomes the new total.

Sets

Open on its own

Set theory in CAT appears as Venn diagram questions in both Quant and Logical Reasoning, often asking for maximum or minimum overlaps. Draw the diagram, fill the innermost region first and keep every region non-negative.

Two sets

FormulaWhen to use
\(n(A \cup B) = n(A) + n(B) - n(A \cap B)\)Number in at least one of the two sets.
Only A \(= n(A) - n(A \cap B)\)Members of A who are not in B, written \(n(A - B)\).
Exactly one \(= n(A) + n(B) - 2\,n(A \cap B)\)Members of one set but not both.
Neither \(= U - n(A \cup B)\)\(U\) is the total number in the universal set.
\((A \cup B)' = A' \cap B'\), \((A \cap B)' = A' \cup B'\)De Morgan's laws for "not A and not B" type statements.

Three sets

FormulaWhen to use
\(n(A \cup B \cup C) = n(A) + n(B) + n(C) - n(A \cap B) - n(B \cap C) - n(A \cap C) + n(A \cap B \cap C)\)Number in at least one of three sets.
Let \(S_1 = n(A) + n(B) + n(C)\), \(S_2 = n(A \cap B) + n(B \cap C) + n(A \cap C)\), \(S_3 = n(A \cap B \cap C)\)Shorthand used in the rows below.
Exactly one \(= S_1 - 2S_2 + 3S_3\)Members of only one of the three sets.
Exactly two \(= S_2 - 3S_3\)Members of exactly two sets, not all three.
At least two \(= S_2 - 2S_3\)Members of two or three sets.
\(S_1 = E_1 + 2E_2 + 3E_3\) and \(n(A \cup B \cup C) = E_1 + E_2 + E_3\)\(E_k\) is the number in exactly \(k\) sets; each person is counted once per set they belong to.
Only A \(= n(A) - n(A \cap B) - n(A \cap C) + n(A \cap B \cap C)\)Members of A and neither of the other two.

Maximum, minimum and subsets

FormulaWhen to use
\(\max(0,\ n(A) + n(B) - U) \le n(A \cap B) \le \min(n(A), n(B))\)Range of the overlap of two sets inside a universe of size \(U\).
\(n(A \cap B \cap C) \ge n(A) + n(B) + n(C) - 2U\)Lower bound on the triple overlap; if the right side is negative, the bound is 0.
Number of subsets \(= 2^n\); proper subsets \(= 2^n - 1\); non-empty proper subsets \(= 2^n - 2\)A set with \(n\) elements, \(n \ge 1\) for the last count.
Venn regions \(= 2^k\) (including the outside region)\(k\) sets in general position: 4 regions for 2 sets, 8 for 3 sets.

Shortcuts and traps

  • Only versus exactly: "only A" means A and no other set; "exactly two" excludes those in all three.
  • At least one is the union: "at least one" equals \(n(A \cup B \cup C)\), not \(S_1\), which counts people more than once.
  • Fill from the centre: write the triple overlap first, then the exactly-two regions, then the only regions.
  • Check every region: when maximising or minimising, a value is valid only if no Venn region becomes negative.
  • Read "neither": if some people are in no set, the union is less than the total; subtract them before using the formulas.