Mental Math

How to Get Faster at Mental Math

Speed isn't talent. It's a different set of procedures than the ones you learned on paper.

Why Mental Math Speed Is a Trained Skill

People who look fast at mental math aren't running your arithmetic faster — they're running a different procedure. School teaches a written method: line up the digits, work right to left, carry the extra into the next column. That method is built for pencil and paper, where a carry can sit quietly in the margin until you need it. In your head, it's clumsy, because you have to hold that carry in memory while also tracking every digit you haven't processed yet.

The techniques below replace the paper method with ones designed for working memory instead of a page. Each one trades a bit of "cleanliness" for a big cut in how much you have to hold in your head at once. None of them require special talent — they're procedures, and procedures get faster with repetition. Learn one at a time and use it for a few days before adding the next; stacking too many new habits at once is what makes people give up on all of them.

Add From the Left, Not the Right

The single highest-leverage habit switch is adding left to right instead of right to left. Right-to-left addition forces you to carry a value into a column you haven't looked at yet — a number you have to store and remember to apply later. Left to right, you handle the biggest place value first and adjust downward, which matches how you already judge size.

Take 47 + 38. Add the tens first: 40 + 30 = 70. Then the ones: 7 + 8 = 15. Combine the two: 70 + 15 = 85. You never had to carry a 1 back into a column you'd already finished — you just added two easy numbers together at the very end.

It scales the same way for bigger numbers. For 267 + 345: hundreds first, 200 + 300 = 500; tens next, 60 + 40 = 100, running total 600; ones last, 7 + 5 = 12, running total 612. Each step is a small addition, and you never hold more than two numbers in your head at once.

Break Numbers Into Friendly Chunks

Some numbers are just easier to work with than others — 100, 50, 25, and multiples of 10 take almost no mental effort. Chunking means temporarily rounding a number to one of these, doing the easy version of the calculation, then correcting for the rounding at the end.

298 + 156 is annoying as written. Round 298 up to 300 — you've now added 2 that wasn't really there — do 300 + 156 = 456, then subtract the 2 back out: 454.

83 − 39 works the same trick in reverse: round 39 up to 40, do 83 − 40 = 43, then add back the 1 you over-subtracted: 44.

The rule that keeps this from going wrong under pressure: round a number up and you'll subtract the difference afterward; round a number down and you'll add it back. Say that correction to yourself the instant you round, not after you've already finished the arithmetic — that's the step people forget when they're moving fast.

Multiplication Shortcuts Worth Memorizing

A handful of specific multipliers come up constantly, and each has a shortcut that turns a two-step multiplication into something closer to a lookup.

None of these are unreliable tricks — they're just algebra (×11 = ×10 + ×1, ×9 = ×10 − ×1, ×5 = ×10 ÷ 2) applied before you start crunching digits instead of after.

Complements: Turn Subtraction Into Addition

Subtraction is where most people lose the most time, because borrowing across a zero breaks the left-to-right flow completely. The fix is to stop subtracting and start adding: reframe the problem around how far the second number is from a round one.

83 − 27: instead of borrowing, subtract 30 (83 − 30 = 53), then add back the 3 you over-subtracted, since 27 is 3 short of 30: 53 + 3 = 56.

For subtracting from a round number, use the 9-complement shortcut. To compute 1000 − 386, subtract each digit of 386 from 9, and the final digit from 10: 9 − 3 = 6, 9 − 8 = 1, 10 − 6 = 4 → 614. This works because 999 − 386 has zero borrowing anywhere (every column is safe on its own), and 1000 − 386 is exactly one more than that: 613 + 1 = 614.

Estimate First, Then Correct

Before calculating anything exactly, get in the habit of estimating the rough size of the answer. It catches errors immediately — if your estimate says "around 1,200" and your exact math comes out to 190, you know a digit got dropped somewhere before you've even committed to the answer.

Estimating can also do real calculation work when you round to a convenient number and correct afterward. 29 × 41: round 29 up to 30 and leave 41 alone — 30 × 41 = 1230. You rounded 29 up by 1, which added one extra, unwanted 41 to the total. Subtract it: 1230 − 41 = 1189.

That's the same rounding-and-correcting logic as chunked addition, just applied to multiplication: pick the nearest easy number, solve the easy version, then adjust by exactly the amount you changed.

The Times-Table Gaps Almost Everyone Has

Most people have solid, instant recall for the 1–5 and 10 times tables, and shaky recall for the middle ground: 6, 7, 8, and 9 combined with each other. Ask most adults for 7 × 8 or 6 × 7 on the spot and there's a pause — those specific pairs get drilled less in school than the "easy" ones, so they never get burned in the way 5 × 5 did.

When one of these won't stick, derive it from a neighbor you already know solidly instead of re-memorizing it in isolation. 7 × 8 is just one more group of 7 than 7 × 7, so 49 + 7 = 56. 6 × 7 is one fewer group of 7 than 7 × 7, so 49 − 7 = 42. Anchoring the shaky facts to the square numbers — which people tend to know cold — closes the gap faster than flashcard repetition of the pair itself.

Distribute Instead of Multiplying Straight

When a multiplication has no obvious shortcut, split one of the factors into a piece you know cold and a piece that's trivial to add back. This is just the distributive property, used on purpose instead of by accident.

14 × 6: instead of tackling 14 × 6 head-on, split 6 into 5 + 1. 14 × 5 = 70 (half of 14 × 10), and 14 × 1 = 14. Add them: 70 + 14 = 84.

The same move works splitting the other factor. 23 × 6: split 23 into 20 + 3. 20 × 6 = 120, and 3 × 6 = 18. Total: 138. Pick whichever factor splits into rounder pieces — that usually means peeling off a multiple of 10 or 5 and handling the small remainder separately.

This generalizes past any two specific numbers: any multiplication can be rewritten as a sum of easier multiplications. The skill isn't memorizing which split to use for which problem — it's recognizing, in under a second, which piece of a number is "free" and which is a fixed small remainder to deal with after.

Turning Technique Into Speed

Reading these techniques doesn't make them automatic. You have to use them under enough time pressure that your brain reaches for the shortcut instead of the school-taught default — that's the gap between understanding a method and actually doing it without thinking, and it only closes with reps.

Speed Calc Duel is built for exactly this: you're racing an opponent to 15 points against a 60-second clock, so there's no time to fall back on right-to-left carrying — either the left-to-right and complement habits are loaded, or you lose ground on every problem. Play Speed Calc Duel when you want raw-speed drilling on the techniques above.

Target Number exercises a different piece of this: given a handful of numbers, you build an expression that lands on an exact target, which means constantly estimating whether a combination is even in the right range before you commit to it. That's the estimate-first habit from earlier, under a ticking clock. Try Target Number to practice sizing up a calculation before you commit to it.

CrossMath Grid flips the direction again — you're filling in missing numbers so a whole grid of equations checks out at once, which rewards the distributive-splitting technique because you're often solving for one piece of a multiplication rather than the whole thing. CrossMath Grid is a good fit once the basics feel comfortable and you want to stress-test them against a clock.

Frequently Asked Questions

How long does it actually take to get noticeably faster at mental math?

Most people feel a real difference within two to three weeks of deliberate practice, as long as you drill one technique at a time instead of trying to absorb all of them at once. Left-to-right addition and the ×11/×9/×5 shortcuts tend to click fastest because they replace a single habit; complements and distributive splitting take a bit longer because they require recognizing when to apply them.

Do I need to memorize times tables all the way up to 12, or is 10 enough?

Solid recall through 10×10 covers the vast majority of everyday mental math, especially once you can derive the shakier facts (6×7, 7×8, and similar) from square numbers you already know. Memorizing 11 and 12 is nice to have but low priority — the ×11 shortcut already handles most of what 11× would give you.

I learned right-to-left addition in school. Is switching to left-to-right really worth relearning?

Yes, if you ever do arithmetic without paper. Right-to-left works fine on a page because a carry can sit in the margin; in your head, that carry is one more thing you have to store while still processing digits you haven't reached. Left-to-right lets you announce a running total as you go, which is both faster and easier to double-check midway through.

What's the fastest way to multiply two two-digit numbers, like 47 × 63?

Combine distribution with rounding: 47 × 63 = 47 × 60 + 47 × 3. 47 × 60 = 2,820 (47 × 6 = 282, then add a zero), and 47 × 3 = 141. Add them: 2,961. There's no single shortcut for arbitrary two-digit pairs — the skill is picking a clean split fast, not recalling a formula.

How do I stop making small errors when I speed up?

Estimate before you calculate. If you know the answer to 29 × 41 should land near 1,200, an exact result of 189 or 11,890 is instantly wrong, even before you re-check the steps. The estimate-first habit catches misplaced decimal points and dropped digits, which are the most common errors people make once they start moving fast.

Are these the same as "Vedic math" tricks?

Some overlap exists — the ×11 shortcut and complement subtraction show up in Vedic math texts too — but this isn't a closed system with its own vocabulary to learn. Every technique here is standard arithmetic (the distributive and commutative properties) applied in an order that's easier to hold in working memory, not a separate method you adopt wholesale.

Play These on Scotix

More Guides

Ready to train?

Jump into a free 1v1 cognitive duel or a solo practice round in seconds. No download required.

Play Scotix Free