PrepShorts · Study sheet · Class 7 Mathematics · Chapter 4, Another Peek Beyond the Point
Chapter 4 · Another Peek Beyond the Point
Estimating first, so a wrong answer is visible
This video could not be loaded. Reload the page to try again.
Sign in with Google10 min.
Keep your place in this chapter — sign in, it’s free.Sign in
One circuit of the Sun takes 365.2422 days and a calendar counts 365. That leftover fifth of a day is why February has a spare one.
The idea
The 0.2422 of a day left hanging at the end of every calendar year is the chapter's demonstration that "too small to matter" is a claim about a scale and not about a number: run it out over a century and it is more than three weeks. The three leap-year rules that follow are not calendar trivia — they are three guesses, and each one is checked the same way, by counting what the scheme delivers, counting what the Earth actually needs, and subtracting. All three missed. The first two missed by enough to be thrown out and replaced; the third is still two-tenths of a day short over a thousand years, and the people who made it knew that, because they had done the subtraction and then decided a residue that size was worth living with. Sizing an answer before you commit to it is the habit the section is teaching, and it is the same habit that lets you put six expressions in order without evaluating a single one.
What you should be able to do
- Explain why a small decimal cannot be dismissed until you know what it is multiplied by
- Compute the accumulated shortfall of a 365-day calendar over 100, 1000 and 10,000 years
- Follow each of the chapter's three leap-year decision trees and classify a given year by it
- Write, as an expression, how many days a stated run of calendar years holds
- Compare what a scheme delivers with what is actually needed, and state whether it overshoots or undershoots
- Rescale a known product or quotient to answer a new one without multiplying again
- Order a set of products and quotients by size using only the position of each factor relative to 1
- Judge whether an answer is plausible before checking it in detail
Words to know
| Term | Definition in one line | First introduced |
|---|---|---|
| leap year | a calendar year given one extra day | printed in §4.4, Part II, pp.88, 91 |
| calendar year | the year as counted by the calendar, as against the Earth's actual circuit | printed in §4.4, Part II, pp.88–92 |
| Gregorian calendar | the calendar the section is describing, named once at the start | printed in §4.4, Part II, p.88 |
| revolution | one complete circuit of our planet round the Sun | printed in §4.4, Part II, pp.88–91 |
| divisible by | the test each decision tree branches on | printed in §4.4, Part II, pp.89–92 |
| adjustment | the book's word for each successive correction to the rule | printed in §4.4, Part II, pp.88, 90, 92 |
| expression | a written formula for a count, evaluated afterwards | printed in §4.4, Part II, pp.89–90 |
| overcompensate | to correct by more than the error, as the four-year rule does | printed in §4.4, Part II, p.89 |
| increasing order | the arrangement the closing exercise asks for | printed in §4.4, "Figure it Out" Q11, Part II, p.95 |
| product | the result of a multiplication, whose size is judged in Q10 and Q11 | printed in §4.4, "Figure it Out", Part II, pp.94–95 |
| quotient | the result of a division, judged the same way | printed in §4.3, Part II, pp.84–87 |
| estimate | working out roughly how big an answer will be before computing it exactly | not printed in this chapter — the section is built on the practice but never uses the word |
| order of magnitude | the explanation's phrase for how many powers of ten separate two answers | an added term, not printed in this chapter |
| sanity check | the explanation's name for the compare-both-totals move the section makes four times | an added phrasing; the book performs the move without labelling it |
Where people slip up
- "A small decimal can be ignored." The section exists to refute this. 0.2422 is small; 0.2422 multiplied by a thousand is 242.2 days of drift — getting on for eight months. Whether a number is negligible depends entirely on what it will be multiplied by.
- "Every fourth year is a leap year." Only under the first of the three schemes. The chapter replaces that rule twice and ends with a three-level test. An explanation that stops at the first tree teaches the rule the chapter spends four pages correcting.
- "The final rule is exact." It is not, and the book says so: over a thousand years the calendar still falls about two-tenths of a day short, and the calendar makers decided to live with it. Presenting the Gregorian rule as perfect throws away the section's conclusion.
- "Estimating is what you do when you cannot be bothered to compute." In this section the estimate is what tells you whether the exact computation is worth doing and whether its answer is credible. Both totals are always computed; the sizing decides what to make of them.
- "Checking means doing the same sum twice." The chapter never re-does a sum. It computes the same quantity by two independent routes — days the scheme delivers, days the Earth needs — and compares. That is what makes an error visible.
- "Q11 needs a calculator." It needs one observation: multiplying by 7.9682 enlarges, multiplying by 0.942368 shrinks slightly, and dividing does the opposite of each. The six expressions can be ordered from that alone.
Ask your teacher a person
Your teacher reads this and writes back, usually within a day. For an instant answer, use Ask the video in the sidebar.
Your class sees the question and the answer. Only your teacher sees that it was you.
No questions on this topic yet.
Worked answers to this chapter’s exercises · this video explains Figure it Out · 4 Q11
Transcript1,449 words
Here is a number, and the only question is whether you may ignore it. Nought point two four two two. One circuit of our planet round the Sun takes three hundred and sixty five point two four two two days. A calendar counts three hundred and sixty five. So every year something is left hanging. Under a quarter of a day, about five and three quarter hours. Small enough that the instinct is to round it away.
Hold on to that instinct. We are about to do the one thing that decides whether it is right. Multiply it by a hundred. Nought point two four two two, a hundred times over, is twenty four point two two. Not hours. Days. Twenty four days adrift in a single century. More than three weeks. Over a thousand years, two hundred and forty two days. Getting on for eight months.
Over ten thousand, two thousand four hundred and twenty two. More than six years out. And notice what did not happen. The number never changed. Only what it got multiplied by. So too small to bother with was never a fact about the number. It is a claim about a scale, and nobody named the scale. Put a marker on the first of January and watch where the planet really is when that date comes round.
After one year it is a quarter of a day behind. Nobody could notice. After a century the calendar runs twenty four days ahead of the sky, and the date has slid nearly a month from its season. Nothing dramatic ever happens. That is the difficulty. It arrives a fifth of a day at a time. So here is a fix, and it is a good one. The leftover is close to a quarter of a day, and four quarters make one.
So every fourth year, add a day. Three hundred and sixty six instead of three hundred and sixty five. Write it as a question with two answers. Is the year divisible by four? One question, two endings. That is the whole rule. Now, and this is the part worth copying, do not admire it. Check it. Check it by counting the same stretch of time two independent ways, and subtracting.
Four years the way the rule counts them. Three ordinary years and a long one. One thousand four hundred and sixty one days. Four years the way the planet counts them. One thousand four hundred and sixty point nine six eight eight. The rule gives more than is needed. Over by nought point nought three one two. Tiny. And by now we know better than to trust the word tiny.
So take a century. Twenty five of its years are divisible by four, so twenty five days go in. Thirty six thousand five hundred and twenty five delivered. Thirty six thousand five hundred and twenty four point two two needed. Over by nought point seven eight. That is the four year error, twenty five times over. The fix for a calendar running slow has made it run fast. Adjust again. The error is small, so the adjustment is small.
Once a century, skip it. A year divisible by a hundred gets no extra day, even though four divides it. That needs a second question, and it goes on top. Divisible by a hundred? If yes, short, and we are done. If no, ask the old question. Two questions, three endings. Count the century again. Twenty four long years instead of twenty five, so one day fewer. And that flips it. Short, by nought point two two of a day.
Plus nought point seven eight became minus nought point two two. The swing is exactly one, because exactly one day came out. One rule is long, the other short. The truth is between them. Before the third attempt, one habit worth stealing. A century holds twenty four long years, and we did not count them one at a time. We wrote it down first. A hundred divided by four, minus a hundred divided by a hundred.
A hundred over four is the years divisible by four. A hundred over a hundred is the one taken back out. Twenty five minus one is twenty four, and seventy six ordinary years are what is left. It gets written before it gets worked out. A tally gives the answer; this gives the reason. Third attempt. We are short, so put something back. Every four hundredth year, give the day back after all.
Three questions now. Divisible by four hundred, then long. If not, divisible by a hundred, then short. If not, divisible by four, then long. And otherwise, short. Three questions, four endings, and every year lands on exactly one of them. Count four hundred years. Four hundred over four, minus four hundred over a hundred, plus four hundred over four hundred. A hundred, minus four, plus one. Ninety seven long years.
Both totals, and subtract. Over by nought point one two of a day. Over that same span the first rule is out by three point one two days, the second by nought point eight eight. So this is the calendar the world runs on. And it is still wrong. Take it to a thousand years and watch. Sort those thousand years by which question catches them first. Two are divisible by four hundred. Eight by a hundred but not four hundred.
Two hundred and forty by four but not a hundred. And seven hundred and fifty by none of them. Two, eight, two hundred and forty, seven hundred and fifty. A thousand. Every year counted once. The long ones are the first group and the third. Two hundred and forty two. Both totals, and subtract. Short, by two tenths of a day. And that leftover was missed by nobody. It is one subtraction, and they almost certainly did it.
Two tenths of a day per thousand years, and they decided it was worth living with. Only somebody who did the arithmetic gets to make that call. Push it further. Ten thousand years. Ten thousand over four, minus ten thousand over a hundred, plus ten thousand over four hundred. Two thousand five hundred, minus a hundred, plus twenty five. Two thousand four hundred and twenty five long years. Both totals, and subtract. Over, by three entire days.
Now stop. Something just happened that is easy to miss. Over a thousand years the calendar was short. Over ten thousand it is long. The sign flipped. The rule repeats every four hundred years. A thousand years is two and a half cycles. It stops halfway through one. Ten thousand is twenty five whole cycles, and each runs over by nought point one two of a day. Twenty five of those is three days. Exactly what we found.
So the thousand year answer was never the real error. It was the rule caught mid stride. One last habit, the same one made small. Suppose you know that seven hundred and fifty six divided by thirty six is twenty one. Then you know a family of answers, and never divide again. Seven hundred and fifty six over three point six. The bottom shrank ten times, so the answer grows ten times. Two hundred and ten.
Seven point five six over nought point three six. Both shrank a hundred times, so the answer does not move at all. Still twenty one. And seventy five point six over nought point nought three six. Top down by ten, bottom down by a thousand. Two thousand one hundred. Every one of those is the same division wearing a different number of zeros. Which is how a wrong answer becomes visible. Twenty one for that last one, and you know the size is wrong.
So finish with six at once. All built on eighteen. Two multiplications, two divisions, and two more with an extra zero. Put them in order, smallest first, without working out a single one. One question per expression, and always the same one. Which side of one is that second number on? Multiplying by something under one shrinks. Multiplying by something above one grows. Dividing does the opposite of both. So multiplying by nought point something lands below eighteen, and dividing by it lands above.
And nought point nought nine is ten times further from one than nought point nine, so it pushes ten times harder. The largest is two hundred, the smallest one point six two. More than a hundredfold apart, and not one sum done. That is the whole habit, and it started with a quarter of a day. Size the answer first, and a wrong one has nowhere to hide.
Where this fits
Taken from the notes each video was made from, not from the reading order — these are the ideas this one rests on and the ones that later rest on it.
Builds on
- Multiply as whole numbers, then count the decimal digitsClass 7 · Ch 4, Another Peek Beyond the Point
- Why multiplying by a decimal below 1 shrinks a numberClass 7 · Ch 4, Another Peek Beyond the Point
- Dividing when the divisor has a decimalClass 7 · Ch 4, Another Peek Beyond the Point
- Writing a situation as an expression before computing itClass 7 · Ch 2, Arithmetic Expressions
Either side of this one
- Divisions that never endClass 7 · Ch 4, Another Peek Beyond the Point
- Turning a vague comparison into a question with a data answerClass 7 · Ch 5, Connecting the Dots...