Topic hub · 32 articles
Mental arithmetic is a set of methods, not a talent. People who calculate quickly are almost never doing the written algorithm faster in their heads — they are using a different method, one chosen to suit the numbers in front of them and to keep as little as possible in working memory at once.
Written methods are optimised for paper: they work right-to-left, they generalise to any size of number, and they require you to store digits until the end. Mental methods work left-to-right, get the magnitude right first, and lean on structure — near-doubles, complements to ten, distributing over a round number. That is why 48 × 5 is easier as half of 480 than as a column multiplication.
These articles cover the methods themselves, when each one applies, the arithmetic facts worth knowing outright, and how to build the fluency that makes the methods usable under time pressure.
Speed is what happens when a good method becomes automatic. Drilling a bad method just makes you fast at the hard way.
Estimation is not a consolation prize for failing to get the exact answer. In most real situations it is the answer you needed, and it catches the errors exact methods produce.
The skill worth building is recognising which shortcut the numbers invite. The operation-specific guides are organised around that recognition.
Mental arithmetic methods with every worked example checked, an honest account of what daily practice does, and a way to drill it in a few minutes.
A list is a memory problem before it is an arithmetic problem. Hold one running total, group to round numbers when you can see the list, and check it twice.
Round every price up, never to the nearest. Biased rounding keeps your running total above the real figure, which is the only property a budget needs.
Most so-called careless arithmetic errors are slips, not gaps in method, and the two need opposite fixes. A taxonomy plus a checking routine that takes seconds.
Four ways to decide which fraction is bigger, ranked by cost: benchmark against a half, match numerators, measure distance from one, then cross-multiply.
Round an awkward number, do the easy sum, then undo the change. The hard part is the undo, and the direction differs between addition and subtraction.
Stop taking away and measure the gap instead. Counting up turns three-digit subtraction into two or three easy hops, and it is how change is made at a till.
A digit-by-digit multiplication that produces the answer right to left with running carries, instead of building partial products and adding them at the end.
Sales tax rates are awkward on purpose. Break the rate into blocks — 8.25% is 10% minus 2% plus a quarter of 1% — and the arithmetic becomes two easy steps.
Reverse a discount by dividing, never by adding the percentage back. Here is why, plus the fraction shortcuts that make the division easy.
One number, three notations. The anchors worth memorising, how to scale a denominator to 100, and the rule that says which fractions terminate as decimals.
What a supportive at-home session with number facts looks like, why the timer should come last, and how a parent's own attitude changes the outcome.
The lattice grid and the crossing-lines trick are the same place-value bookkeeping in two costumes. Here is how each works, and where both stop being useful.
Add the big parts first, and a written-method carry turns into a running-total correction. The full method, the failure case, and how to absorb it.
What researchers mean by mathematics anxiety, the working-memory account of why it affects performance, and what a practice app can honestly offer.
Five is ten halved and twenty-five is a hundred quartered, so the multiplication becomes a shift and a division. Pick which number to divide first.
Nine is ten minus one, so multiply by the round number and take one lot away. The same move runs at 99, 999 and 9,999, with a digit-sum check built in.
Complements to 10, 100 and 1000 are the layer every mental addition method assumes: what they are, the digit rule with one trap, and how adults install them.
Most proportion problems are scaling problems. Find the factor, apply it, and keep cross-multiplication for the cases where no friendly factor exists.
Halve one number, double the other, discard remainders, then add the rows where the halved side is odd. It works because that column is binary.
Reduce in a fixed order — halve, clear the 5s, then use digit sums for 3 and 9 — and you cut the numbers down before the expensive tests start.
Memorise the squares to 30 and the cubes to 15, then use a last-digit filter and a digital-root check to read exact roots straight back off them.
Add the same amount to both numbers and the difference does not change, but the borrow disappears. Here is why that is safe, and how to pick what to add.
Recall replaces computation, so the capacity spent working out 7 times 8 goes to the rest of the problem. That is the mechanism, and its limit.
You rarely need the price per unit. Scale the smaller pack up to the larger one and compare two totals — one multiplication instead of two awkward divisions.
Slow mental arithmetic has several distinct causes, and each needs a different fix. Here is how to tell them apart from your own accuracy and timing record.
Every divisibility rule is a statement about what powers of ten leave behind. Here is the reasoning for 2, 3, 4, 5, 6, 8, 9, 10 and 11, worked out.
Halving and doubling, factor splitting, divisibility checks and remainder handling: the division methods that survive contact with a real number.
Rounding, front-end estimation, compatible numbers and order-of-magnitude checks — the techniques that make arithmetic useful when precision is not the point.
Doubling and halving, multiplying by 11, the difference of squares, and near-round-number tricks — each with the algebra that makes it work.
The 1% and 10% anchors, the commutative trick that makes hard percentages easy, tips and discounts, and how to reason about percentage change correctly.
Base-100 multiplication, percentage reversal, unit conversion and root estimation — every method worked through, with the case where each one fails.