Mathematical Modelling (Appendix)

An adult human body contains roughly 1,50,000 km of arteries and veins. The human heart pumps 5 to 6 litres of blood every 60 seconds. The surface of the Sun sits at about 6,000°C. Nobody measured these numbers by unwinding a corpse's blood vessels, draining a human body litre by litre, or flying a thermometer to the Sun. Every one of these figures came from mathematical modelling — the craft of describing a real situation in the language of mathematics closely enough that solving the mathematics tells you something true about the world. This appendix revisits an idea you first met in Class IX and gives it a disciplined shape: five stages that turn a fuzzy real-life question into a solvable equation and then turn the answer back into a sensible real-life statement. You will estimate a hidden population of fish or parrots by marking a sample and recapturing it. You will uncover the true, hidden rate of interest buried inside a shopkeeper's instalment scheme, and discover that the "better deal" is not always the one that looks cheaper at first glance. You will use probability to decide what number to guess in a dice game. And you will meet Fibonacci's rabbits, whose population doubles almost every month and whose growth ratio settles, uncannily, on the golden ratio. None of this is examined the way a circle theorem or a quadratic formula is examined — it is enrichment, a chance to see how the mathematics you already know (ratio and proportion, simple interest, probability, sequences) becomes a working tool the moment real numbers and real assumptions are attached to it. The five-stage habit you build here — understand, formulate, solve, interpret, validate — is exactly the habit engineers, epidemiologists, economists, and data scientists use every day, which is precisely why NCERT closes the Class 10 course with it.

Key takeaways

A mathematical model is a mathematical description of a real-life situation; mathematical modelling is the process of building that model, solving it, and interpreting the result back in real terms.
Modelling proceeds in five stages: understanding the problem, mathematical description and formulation, solving the mathematical problem, interpreting the solution, and validating the model against reality.
Simplifying assumptions are unavoidable — every model trades some realism for solvability, so there is no perfect model, only good ones and better ones.
The capture-recapture method estimates a hidden population (fish in a lake, parrots in a field) from a single ratio-and-proportion equation built on marking and remixing a sample.
Instalment schemes hide an effective rate of interest that modelling can uncover, month by month, using the ordinary simple-interest formula on the shrinking outstanding balance.
The dice-sum guessing game shows probability itself acting as a model: assign each outcome a likelihood from the 36 equally likely pairs, and let the highest-probability sum, 7, guide the decision.
When a model's prediction does not match reality, the fix is not to abandon modelling but to revisit the Step 1 assumptions and rebuild — validation can loop back as many times as needed.
Mathematical modelling is deliberately interdisciplinary: it underlies weather forecasting, drug-trial design, nuclear-reactor safety, market-demand forecasting, school-enrolment planning, and election exit polls.
This appendix sits beyond the Class 10 board syllabus's core marks weightage, but the five-stage habit of thought reappears in every quantitative field, from engineering to economics to public health.
Key idea: A mathematical model turns a messy real-life question into a solvable equation through five stages — understand, formulate, solve, interpret, validate — and it is judged not by elegance but by how well its answer survives contact with reality.

The core insight

Mathematical modelling converts a real-life problem into a mathematical one by simplifying it with stated assumptions, solving the resulting equations, interpreting the solution back in real terms, and validating it against what actually happens — looping back to revise the assumptions whenever the model's prediction fails to match reality.

What this chapter covers

This appendix has four moving parts, and the article follows them in order. First, what a mathematical model is — the definition, and NCERT's own list of situations (from the mass of the Earth to the arrival of the monsoon) where modelling is the only realistic route to an answer. Second, the five-stage cycle that every model follows, worked through completely on the classic fish-in-a-lake population problem, and pictured as the flowchart NCERT calls Fig. A2.1. Third, two fully worked illustrations taken directly from the chapter — the dice-sum guessing game (where probability itself is the model) and the instalment-scheme bicycle problem (where simple interest hides an effective rate that must be dug out algebraically) — followed by three step-by-step derivations that generalise these methods so you can apply them to any similar numbers. Fourth, why modelling matters at all: the three broad reasons NCERT gives — to understand, to predict or forecast, and to estimate — with real examples from marketing, education planning, and elections. The worked examples section then applies the capture-recapture and instalment-scheme methods to fresh numbers from NCERT's own exercises, and a higher-order-thinking example pushes the Fibonacci rabbit problem to its logical, slightly startling conclusion.

The core ideas

What is a mathematical model?

A mathematical model is a mathematical description of some real-life situation. Mathematical modelling is the process of creating that model, solving it, and using the solution to understand or act on the real-life problem it came from. In short: take a real-world problem, convert it into an equivalent mathematical problem, solve the mathematical problem, and interpret its solution back in the situation of the real-world problem — then check that the solution actually makes sense. NCERT lists eleven situations where modelling is indispensable because direct measurement is impossible, impractical, or absurdly expensive: the width and depth of a river at an unreachable place; the mass of the Earth and other planets; the distance between Earth and another planet; the arrival of the monsoon; the trend of the stock market; the volume of blood inside a person's body; a city's population ten years from now; the number of leaves on a tree; the parts-per-million of pollutants in a city's air and their effect on the environment; and the temperature at the Sun's surface. The common thread: direct measurement is either impossible (no thermometer reaches the Sun), destructive (you cannot drain a living person's blood), or premature (a population ten years from now has not happened yet). Modelling is how mathematics answers questions observation alone cannot.

The five stages of mathematical modelling

Every mathematical model, however different the numbers, moves through the same five stages.

Step 1 — Understanding the problem. State the real problem plainly, and simplify it by making assumptions that strip away detail you cannot handle mathematically. If the problem is "how many fish are in this lake," you accept immediately that you cannot capture and count every fish; you must work from a sample.

Step 2 — Mathematical description and formulation. Translate the simplified problem into mathematical language: define variables, write equations or inequalities, gather data into tables, draw graphs, or calculate probabilities. For the fish problem, you decide to mark a sample, release it back into the lake, let it mix, then draw a second sample and see what fraction of the second sample is marked.

Step 3 — Solving the mathematical problem. Apply ordinary mathematical technique — here, ratio and proportion — to the formulation from Step 2. Suppose 20 fish are caught, marked, and released, and a later sample of 50 fish contains 5 marked ones. Then $\tfrac{5}{50} = \tfrac{1}{10}$ of the population is marked, and since the 20 marked fish represent that same fraction of the total, $\tfrac{1}{10}$ of the population equals 20, so the whole population is $20 \times 10 = 200$ .

Step 4 — Interpreting the solution. Read the mathematical answer back in real terms: "the lake contains approximately 200 fish."

Step 5 — Validating the model. Return to the real situation and ask whether 200 is believable. If new evidence (say, repeating the sampling gives wildly different answers) suggests the model is off, you revise the Step 1 assumptions — perhaps the marked fish did not mix uniformly, or the sample was too small — and rebuild. The five stages loop rather than run in a straight line: a validation failure sends you back to revise assumptions and repeat the formulation and solving stages with a better model.

Figure

A flowchart of the five stages of mathematical modelling starting from a real life problem, through simplifying assumptions, mathematical formulation, solving, and interpreting, looping back through changed assumptions whenever validation fails, and ending at a suitable model when validation succeeds

Illustration: probability as a model — guessing the sum of two dice

Probability itself can be the mathematical model. Suppose a teacher will throw a pair of dice, and before the throw you must guess the sum that will show. A correct guess earns two points, a wrong one loses two. What number should you guess?

Step 1 (Understanding). You need to know which sums are most likely to occur.

Step 2 (Formulation). Model a throw of two dice as a random choice from the 36 equally likely ordered pairs $(1,1), (1,2), \ldots, (6,6)$ , the first entry being the first die and the second entry the second die.

Step 3 (Solving). Summing each pair gives possible totals of 2 through 12. Counting how many of the 36 pairs give each total produces the probability table: $P(2)=\tfrac{1}{36}$ , $P(3)=\tfrac{2}{36}$ , $P(4)=\tfrac{3}{36}$ , $P(5)=\tfrac{4}{36}$ , $P(6)=\tfrac{5}{36}$ , $P(7)=\tfrac{6}{36}$ , $P(8)=\tfrac{5}{36}$ , $P(9)=\tfrac{4}{36}$ , $P(10)=\tfrac{3}{36}$ , $P(11)=\tfrac{2}{36}$ , $P(12)=\tfrac{1}{36}$ .

Step 4 (Interpreting). The sum 7 has the highest probability, $\tfrac{6}{36}=\tfrac16$ , so 7 is the best repeated guess.

Step 5 (Validating). Throw the dice many times, build a relative-frequency table, and check it against the theoretical probabilities. A large mismatch would suggest the dice are biased, and the model would need new data to correct it.

Figure

A bar chart showing the probability of each possible sum from two thrown dice, rising from a small probability at sum two to a peak of six thirty sixths at sum seven and falling symmetrically back down to sum twelve

Illustration: uncovering the true cost of an instalment scheme

Not every model needs a graph or a sample; some need only careful bookkeeping. When a customer cannot pay the full cash price of an article at once, a trader offers an instalment scheme: a cash down payment at purchase, followed by the balance paid in instalments. Because the seller is not receiving the full amount immediately, the deferred balance carries interest — usually simple interest, when the whole balance is cleared within a year. Historically, charging interest on borrowed money was sometimes disguised (borrowing in one currency and repaying in another let people evade laws against interest); today it is stated openly, but the effective rate a scheme charges is often hidden inside instalment amounts rather than announced directly. Modelling lets you extract that hidden rate and compare it honestly against a bank's or a financial society's quoted rate, which is exactly what NCERT's Juhi-and-the-bicycle example does, and what the derivation and worked examples below generalise.

Why mathematical modelling matters

Mathematical modelling is deliberately interdisciplinary — mathematicians and specialists in other fields pool expertise to improve products, design new ones, or predict behaviour. Three broad reasons explain its importance.

To gain understanding. A model that reflects the essential behaviour of a real system lets you analyse that system, and building the model itself reveals which factors matter most and how the parts of the system relate to each other.

To predict, forecast, or simulate. Often it is expensive, impractical, or impossible to experiment directly with a real system — weather prediction, drug efficacy trials in humans, and the optimal design of a nuclear reactor are all cases where modelling substitutes for direct experiment. Forecasting matters to organisations too: marketing departments need reliable demand forecasts, and a school board needs to forecast enrolment before deciding where to open new schools. Most forecasting techniques analyse past data to find a pattern, then extend it into the future.

To estimate. Estimating large values — the trees in a forest, the fish in a lake — is one of modelling's oldest jobs. Before an election, contesting parties want to estimate how many voters in a constituency will support them, using that estimate to shape campaign strategy; exit polls are a widely used tool for predicting the number of seats a party will win.

Derivation AThe capture-recapture method for estimating a hidden population

The goal here is to turn "how many fish live in this lake" — a question you cannot answer by counting — into a single algebraic equation solvable from two small samples. The method rests on one assumption that must be stated and later checked: once released, marked individuals mix back into the population uniformly, so any later sample contains marked and unmarked individuals in roughly the same proportion as the population as a whole.

Step 1. Capture a first sample of size $n_1$ , mark every individual in it, and release them back into the population to mix thoroughly with the unmarked remainder.

Step 2. After mixing, capture a second, independent sample of size $n_2$ , and count how many carry a mark; call this count $m$ . The marked fraction of the second sample is $\dfrac{m}{n_2}$ .

Step 3. State the assumption: this fraction represents the marked fraction of the whole population, whose unknown size is $N$ . Since exactly $n_1$ individuals in the population are marked, that fraction is also $\dfrac{n_1}{N}$ . Setting the two equal gives $\dfrac{m}{n_2} = \dfrac{n_1}{N}$ .

Step 4. Cross-multiply to solve: $mN = n_1 n_2$ , so $N = \dfrac{n_1 n_2}{m}$ — the estimated population is the product of the two sample sizes divided by the number recaptured marked.

Step 5. Apply it to NCERT's lake example: $n_1 = 20$ marked and released, $n_2 = 50$ drawn later, $m = 5$ found marked. Then $N = \dfrac{20 \times 50}{5} = 200$ , matching the five-stage worked example above. The formula is only as good as its mixing assumption; if marked individuals are easier or harder to recapture, $N$ will be biased — exactly the failure validation is meant to catch. Verified: with $n_1=20$ , $n_2=50$ , $m=5$ : $N = (20 \times 50)/5 = 200$ , matching NCERT's Step 3 result exactly.

Go deeper · Derivation B · 1 minThe general rate-of-interest equation for an instalment scheme

Derivation BThe general rate-of-interest equation for an instalment scheme

The goal is to derive a single formula for the effective annual simple-interest rate $r$ that a trader is actually charging inside an instalment scheme with a cash down payment and $k$ equal monthly instalments, so this hidden rate can be compared honestly against a bank's quoted rate. The method treats the deferred balance as a sequence of shrinking one-month loans, each charged simple interest for one month, summed across all $k$ months.

Step 1. Let $C$ be the cash price and $D$ the cash down payment. The balance owed is $B = C - D$ , cleared in $k$ equal monthly instalments of amount $I$ each.

Step 2. The total paid across all instalments is $kI$ , so the total interest paid is $kI - B$ — the extra money handed over beyond the balance actually owed.

Step 3. Track the outstanding principal month by month: $B$ in month 1, $B-I$ in month 2, $B-2I$ in month 3, down to $B-(k-1)I$ in the final month. Summing across all $k$ months gives $P_{\text{total}} = kB - I\bigl(0+1+\cdots+(k-1)\bigr) = kB - I \cdot \dfrac{k(k-1)}{2}$ .

Step 4. Model the interest as simple interest at annual rate $r$ on this running principal, one month ( $\tfrac{1}{12}$ year) at a time: interest $= P_{\text{total}} \times \dfrac{r}{100} \times \dfrac{1}{12}$ . Setting this equal to $kI - B$ from Step 2 and solving gives $r = \dfrac{1200\,(kI - B)}{P_{\text{total}}}$ .

Step 5. Check against NCERT's bicycle example: $B = \text{₹}1200$ , $I = \text{₹}610$ , $k = 2$ . Then $P_{\text{total}} = 2(1200) - 610(1) = 1790$ , interest $= 1220 - 1200 = 20$ , so $r = \dfrac{1200 \times 20}{1790} \approx 13.41\%$ per annum — well above the bank's 10%, confirming Juhi should borrow from the bank instead. Verified: $B=1200$ , $I=610$ , $k=2 \Rightarrow P_{\text{total}} = 1790$ , interest $= 20$ , $r \approx 13.41\%$ , consistent with NCERT's own equation $1790 \times r \times 1/1200 = 20$ .

Go deeper · Derivation C · 1 minThe Fibonacci recurrence as a model of rabbit population growth

Derivation CThe Fibonacci recurrence as a model of rabbit population growth

The goal is to derive, from the stated biological assumptions alone, the recurrence relation that governs Fibonacci's rabbit problem, showing that the number of pairs in any month is exactly the sum of the pairs in the two preceding months — not a pattern spotted in a table, but a direct consequence of the assumptions.

Step 1. State the assumptions as the formulation: one newborn pair exists in month 0; every pair becomes fertile at age 2 months and thereafter produces exactly one new pair every month; no pair ever dies. Let $F(n)$ denote the pairs alive in month $n$ .

Step 2. Split the pairs alive in month $n$ into two groups: those already alive in month $n-1$ (all survive, since nobody dies), and the new pairs born during month $n$ . A pair can only give birth once it is at least 2 months old, meaning it must have existed already in month $n-2$ ; every such pair is now fertile and produces exactly one new pair. So the number of new pairs equals $F(n-2)$ .

Step 3. Combine the two groups: $F(n) = F(n-1) + F(n-2)$ , with starting values $F(0) = F(1) = 1$ . This is the Fibonacci recurrence, and it is not empirical — it is an algebraic consequence of "nobody dies, every pair matures in 2 months, every fertile pair adds one pair per month."

Step 4. Confirm against NCERT's own table: $F(2) = 1+1 = 2$ , $F(3) = 2+1 = 3$ , $F(4) = 3+2 = 5$ , and so on up to $F(16) = 1597$ , matching NCERT's table for months 0 through 16 exactly. Verified: recomputing $F(n) = F(n-1)+F(n-2)$ from $F(0)=F(1)=1$ up to $n=16$ reproduces 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597 term for term against NCERT's table.

Worked examples

Board level - estimating a hidden parrot population by capture-recapture (extends NCERT Exercise A2.2, Q1)

An ornithologist wants to estimate the number of parrots in a large field. She nets and rings 32 parrots and releases them. The following week she nets 40 parrots, of which 8 are found to be ringed. (i) What fraction of her second catch is ringed? (ii) Estimate the total number of parrots in the field.

Step 1. Understanding the problem: the unknown is the total parrot population $N$ in the field, and the tool available is a mark-and-recapture sample, exactly the situation modelled in the first derivation above.

Step 2. Formulate: here $n_1 = 32$ (ringed and released), $n_2 = 40$ (second catch), and $m = 8$ (ringed birds found in the second catch). The fraction of the second catch that is ringed is $\dfrac{m}{n_2} = \dfrac{8}{40} = \dfrac{1}{5}$ .

Step 3. Solve using $N = \dfrac{n_1 n_2}{m}$ from the capture-recapture derivation: $N = \dfrac{32 \times 40}{8} = \dfrac{1280}{8} = 160$ .

Step 4. Interpret: the field is estimated to hold about 160 parrots.

Step 5. Validate: the estimate assumes the ringed parrots mixed uniformly back into the flock and that no parrots entered or left the field between the two catches; repeating the exercise with a fresh pair of samples and averaging the results would tighten the estimate. Answer: the ringed fraction of the second catch is $\tfrac{1}{5}$ , and the estimated parrot population is 160.

Board level - comparing an instalment scheme against a loan (extends NCERT Exercise A2.2, Q3)

A TV can be bought for ₹24000 cash, or for ₹8000 cash down payment plus six monthly instalments of ₹2800 each. Ali has ₹8000 and is deciding between the instalment scheme and borrowing the balance from a financial society that charges 18% per annum simple interest. Which option is more economical?

Step 1. Understanding: Ali needs to compare the effective interest rate hidden in the instalment scheme against the society's stated 18%.

Step 2. Formulate using the general instalment formula: cash price $C = 24000$ , down payment $D = 8000$ , so the balance is $B = C - D = 16000$ ; the scheme has $k = 6$ monthly instalments of $I = 2800$ each.

Step 3. Solve. Total paid in instalments is $kI = 6 \times 2800 = 16800$ , so the total interest paid is $kI - B = 16800 - 16000 = 800$ . The principal outstanding across the six months is $P_{\text{total}} = kB - I\cdot\dfrac{k(k-1)}{2} = 6(16000) - 2800 \times \dfrac{6 \times 5}{2} = 96000 - 2800 \times 15 = 96000 - 42000 = 54000$ . Then $r = \dfrac{1200 (kI - B)}{P_{\text{total}}} = \dfrac{1200 \times 800}{54000} = \dfrac{960000}{54000} \approx 17.78\%$ per annum.

Step 4. Interpret: the shopkeeper's scheme charges roughly 17.78% per annum, which is very slightly below the financial society's 18%.

Step 5. Validate: the two figures are close enough that other real-world costs (stamp duty or paperwork on a bank loan, or the inconvenience of monthly visits under the instalment scheme) could tip the decision either way; purely on the interest rate, the instalment scheme is marginally cheaper. Answer: the instalment scheme's effective rate is about 17.78% per annum, a shade below the society's 18%, so Ali should prefer the instalment scheme.

Higher-order thinking (HOTS) - the growth rate of Fibonacci's rabbits (extends NCERT Exercise A2.1, Q1)

Using the Fibonacci recurrence $F(n) = F(n-1) + F(n-2)$ with $F(0) = F(1) = 1$ , examine how the population's month-to-month growth factor $F(n+1)/F(n)$ behaves as $n$ grows, and assess whether the model remains believable indefinitely.

Step 1. Understanding: NCERT's table already lists $F(n)$ from $n=0$ to $16$ , ending at $F(16) = 1597$ pairs after just 16 months. The higher-order question is not just "how many pairs" but "at what steady rate is the population multiplying, and can that rate keep applying forever."

Step 2. Formulate the growth factor $\rho(n) = \dfrac{F(n+1)}{F(n)}$ and track it as $n$ increases.

Step 3. Solve by computing $\rho(n)$ for several late values of $n$ : $\rho(9) = \dfrac{F(10)}{F(9)} = \dfrac{89}{55} = 1.6182$ , $\rho(12) = \dfrac{F(13)}{F(12)} = \dfrac{377}{233} = 1.6180$ , $\rho(14) = \dfrac{F(15)}{F(14)} = \dfrac{987}{610} = 1.6180$ , $\rho(15) = \dfrac{F(16)}{F(15)} = \dfrac{1597}{987} = 1.6180$ . The ratio is converging, to four decimal places, on $1.6180$ — the golden ratio, $\varphi = \dfrac{1+\sqrt5}{2} \approx 1.618$ .

Step 4. Interpret: after roughly ten months the rabbit population is not just growing, it is growing at an almost perfectly constant monthly multiplication factor of about $1.618$ , so the growth is effectively exponential with base $\varphi$ . This is a striking and non-obvious consequence of the simple recurrence — the golden ratio was never assumed, it emerged from "no deaths, mature at 2 months, one new pair a month."

Step 5. Validate — the higher-order-thinking crux. A factor above 1.6 every month is explosive: continuing the recurrence would push the population past a million pairs within about 30 months. No real rabbit population sustains that; food, space, predators, and mortality all act against unlimited growth, none of which the model includes. The Fibonacci model is an excellent short-run model because it captures the reproduction mechanism honestly, but a poor long-run one because it omits every real population's eventual resource limit — a complete model would need a carrying-capacity term (logistic growth), beyond Class 10 mathematics. Answer: the growth factor converges to the golden ratio $\varphi \approx 1.618$ ; the exponential model is trustworthy for roughly the first year but becomes unrealistic beyond it. Timing note: budget about 6 to 8 minutes for the full ratio computation and the validation discussion — the arithmetic in Step 3 is quick, but the value of this problem is in Step 5's honest critique of the model's own limits, which is exactly what "validating the model" is meant to train you to do.

Common traps

Treating the mathematical answer as automatically true. Solving the equation is not the end of modelling — the answer must still be interpreted and checked against reality. A population estimate of 200 fish is a claim about the world, not a proven fact.
Forgetting to state assumptions explicitly. Every model here rests on assumptions (uniform mixing, no deaths, constant monthly instalments) that must be named, because they are exactly what you revisit if the model fails.
Assuming a more complicated model is automatically better. NCERT is explicit: there is no perfect model, only good ones and better ones, balancing simplicity against accuracy.
Confusing total money paid with the interest-rate comparison. In instalment problems, "total paid" is not the right basis for comparison unless both options run the same length of time; the honest comparison is the effective annual rate.
Extrapolating a short-run pattern indefinitely. A recurrence that fits perfectly for 16 months says nothing about month 160, because the assumptions that generated it stop being realistic long before then.

Common trapThe biggest misconception: solving the equation is the whole job

Students who have just learned ratio and proportion, simple interest, or a recurrence relation often treat mathematical modelling as nothing more than "set up the right formula and compute." That is only Steps 2 and 3 of a five-stage process. The real discipline lives in Step 1 (stating out loud which simplifying assumptions you are making) and Step 5 (checking whether the answer holds up against reality, and being willing to rebuild if it does not). NCERT says it plainly: "there are no perfect models, there are good ones and yet better ones." A numerically correct solution to the wrong assumptions is not a good model — it is a wrong answer with confident arithmetic. Always finish the cycle: interpret, then validate, then be ready to revise.

You should be able to…

State the definitions of a mathematical model and mathematical modelling, in your own words, with an example.
List, from memory, at least four of NCERT's eleven real-life situations where modelling is the only practical route to an answer.
Name and explain, in order, all five stages of the modelling cycle: understanding, formulation, solving, interpreting, validating.
Explain why validation can send you back to revise Step 1's assumptions, and why there is no such thing as a perfect model.
Apply the capture-recapture formula $N = n_1 n_2 / m$ to estimate a hidden population from two samples.
Compute the probability distribution for the sum of two dice, and identify why 7 is the best single guess.
Set up and solve an instalment-scheme problem to find the effective annual rate of interest, and compare it against a stated bank or society rate.
Derive the Fibonacci recurrence $F(n) = F(n-1) + F(n-2)$ from its underlying reproduction assumptions, rather than only recognising the pattern in a table.
Recognise the three broad reasons modelling matters: to understand a system, to predict or forecast it, and to estimate an otherwise-unmeasurable quantity.

Ask the tutor: if you want to work through a capture-recapture or instalment-scheme problem with different numbers, or see how the Fibonacci growth ratio is proved to converge exactly to the golden ratio, ask the tutor to walk through it step by step.

Is mathematical modelling going to be tested in my Class 10 board exam?

Generally no, not as a separate, heavily weighted topic the way quadratic equations or trigonometry are. This appendix is explicitly framed by NCERT as enrichment — a way to see how the mathematics you already know (ratio and proportion, simple interest, probability, sequences) becomes a working tool on real problems. Some boards include a short conceptual question on the stages of modelling or a simple capture-recapture or instalment-scheme numerical, so check your own board's latest sample paper, but do not expect it to carry the same marks weightage as the core chapters. Treat it as the chapter that shows you why the algebra you spent the year learning is actually useful.

Why do we need assumptions at all — why not just solve the problem exactly?

Because most real-life problems are too complicated to solve exactly with the mathematics you have. You cannot count every fish in a lake or every leaf on a tree. Assumptions are how you trade some realism for solvability — you assume marked fish mix uniformly, precisely so the resulting mathematics becomes tractable with ratio and proportion. The skill is not avoiding assumptions, it is choosing ones simple enough to solve but close enough to reality to be useful — and being ready to revise them if validation shows they were too crude.

How is capture-recapture different from just guessing or averaging?

Capture-recapture is not a guess; it rests on a precise, checkable assumption — that marked individuals, once released, mix back into the population so thoroughly that any later sample reflects the same marked-to-total proportion as the whole population. Given that, the estimate $N = n_1 n_2 / m$ follows from ratio and proportion, not intuition. What makes it a genuine model is that the assumption is testable: repeat the second sample and get wildly different marked fractions each time, and that is direct evidence the mixing assumption has failed, calling for revision.

The instalment-scheme problems look just like simple-interest problems I already know. What is actually new here?

The arithmetic tool — simple interest, $I = P \times r \times t / 100$ — is one you already know. What is new is the modelling layer: recognising that an instalment scheme is a sequence of shrinking one-month loans, each charged interest separately, so you must sum the outstanding principal month by month before solving for the single annual rate $r$ that explains all of it. Turning "cash price, down payment, and monthly instalments" into one equation is the formulation stage; comparing that rate honestly against a real alternative like a bank loan is the validation.