GMAT® Computation: When to Do the Math (and When to Let Go)

Here's something that happens to a lot of people preparing for the GMAT® Focus Edition.

You see a computation problem. You recognize the type. You feel 90% confident you know the answer. So you move quickly, pick the answer that looks right, and move on.

Then you find out you were wrong.

It's one of the most common patterns we see at The GMAT® Strategy (TGS) — and after years of coaching students through this exact scenario, we can tell you: it's almost never a knowledge problem.

The math was well within your abilities. The issue was something different — a dropped decimal place, a miscounted digit, a step skipped because it felt unnecessary.

That's the core challenge with computation on the GMAT®. The material isn't hard. The execution is.

We walked through three real computation problems in Episode 46 of Real GMAT® Problems, our podcast series. Here's what they revealed about how to do computation well — and when to make the tough call to let a problem go.

Why Does the GMAT® Care So Much About Execution?

Most standardized tests reward accuracy. Get the right answer, earn the point.

The GMAT® does too, but with an important twist: the difficulty of questions you see affects your score. The exam is adaptive. If you're getting easy questions right, it shows you harder questions. If you start missing easy questions, it drops you back down.

This means missing a problem you knew how to do has an outsized negative effect. You don't just lose the point — you lose the positioning that comes from demonstrating you can handle that level of difficulty.

The practical implication: on familiar problems, slow down. Do the full math. Going from 90% confident to 100% certain is worth the extra 30 seconds, because the cost of that 10% margin of error is much higher than the cost of the time spent eliminating it.

That sounds counterintuitive. We're wired to speed up on familiar material. But the GMAT® rewards the opposite habit.

The First Problem: 0.1 + 0.1² + 0.1³

This is the warm-up problem from Episode 46. Decimal exponents. Quick to compute, easy to verify.

Some people can look at this and know instantly that the answer is 0.111. If you're one of them, great. But if there's any doubt — even a small one — the right move is to work through it.

Here's the full approach for 0.1²: write 0.1 × 0.1, ignore the decimal points, do 1 × 1 = 1, then count total decimal places (two), and shift the decimal two places left. Result: 0.01.

For 0.1³: take 0.01 × 0.1, do 1 × 1 = 1, count three decimal places total, shift three places left. Result: 0.001.

Then add:

0.100 + 0.010 + 0.001 = 0.111

The answer is (B). Full solution: 0.1 + 0.1² + 0.1³ — Worked Solution.

The lesson isn't the math. The lesson is why you should do it even when you think you don't need to. The GMAT® scoring algorithm means every "easy" question you drop is more expensive than it looks.

The Second Problem: Converting Decimals to Fractions

The second problem — (0.045 × 1.9) / (0.03 × 0.005 × 0.1) — looks intimidating at first glance. It's a fraction full of decimals. There are a lot of places to go wrong.

The cleanest path through problems like this is also the simplest: convert every decimal to a fraction, then flip the denominator fractions and multiply everything across.

0.045 becomes 45/1,000. 1.9 becomes 19/10. 0.03 becomes 3/100. And so on.

Once everything is in fraction form, you're dividing fractions — which means multiplying by the reciprocal. Flip the bottom fractions and multiply straight across. Then cancel zeros. The computation collapses quickly.

In this case, you wind up with 3 × 19 × 100 = 5,700. The answer is (A).

The broader principle: one approach works for almost every decimal computation problem on the GMAT®. You don't need to memorize a dozen specific decimal tricks. You need one reliable system — convert to fractions, flip, multiply, cancel — and you can apply it across a huge range of problems.

If you're weeks out from test day, building your decimal computation skills gives you more flexibility. If you're days out and feeling unsteady, the fraction conversion approach is a defensible fallback for almost anything they'll throw at you.

The Third Problem: When to Do the Math, and When to Let Go

The third problem asks how many distinct non-zero digits x has, where x = 1/(2² × 3² × 4² × 5²) expressed as a decimal. This is the hard one. About 60% more people miss it than the previous problem.

The solution path exists, but it requires patience. Full worked solution: 1/(2² × 3² × 4² × 5²) — Distinct Non-Zero Digits.

Here's the key move: pull out 2² × 5² = 100. Dividing by 100 only shifts the decimal point — it doesn't introduce any new non-zero digits. So you can set that aside and work with 1/(3² × 4²) = 1/144.

Then long-divide 1 by 144. It takes some work, but the decimal comes out to 0.006944444...

Non-zero digits: 6, 9, 4. Three distinct values. The answer is (C).

But here's the other point this problem teaches.

If you reach a problem during the actual exam and find yourself losing confidence — you started at 80% and now you're at 30% — that feeling is important information. Problems that drain your confidence and time budget at that rate are often the ones worth letting go.

The GMAT® is engineered so you'll miss a significant number of questions. The adaptive algorithm gives you harder and harder problems until you start missing them. That's how it measures where your ceiling is.

Accepting that reality — and acting on it by cutting losses on hard problems quickly — frees up time for the problems you can get right. That's not giving up. That's resource management. The same principle shows up in any leadership context: the hardest decisions aren't the technical ones. They're the ones that require you to let go of something you want to succeed.

What Should You Practice This Week?

Build the habit of doing the full computation. Pick 10 computation problems from the Official Guide. On every problem where you feel 80% or more confident, do the full math anyway. Check your result. Notice how often you would have been right — and how often that extra step would have caught an error you'd otherwise have carried forward.

Practice the fraction conversion approach. Take 5 decimal computation problems and convert every number to a fraction before touching the math. See how quickly the zeros cancel and the computation simplifies. Do this until the conversion feels automatic.

Build your benchmarks. Long division gets faster when you know useful products. 6 × 144 = 864. 9 × 144 = 1,296. Having a few of these memorized can save meaningful time on hard problems.

Practice the letting-go decision. During practice sets, track which problems you spent more than 90 seconds on without clear progress. After the set, note what you would have scored if you had guessed and moved on after 60 seconds. Compare. That comparison teaches you a lot about where to draw the line.

We walked through all three problems in Episode 46 of Real GMAT® Problems:

• Problem 1: 0.1 + 0.1² + 0.1³ — decimal exponents, the value of doing the full computation
• Problem 2: (0.045 × 1.9) / (0.03 × 0.005 × 0.1) — one fraction approach for almost every decimal problem
• Problem 3: x = 1/(2² × 3² × 4² × 5²), how many distinct non-zero digits? — knowing when to do the math, and when to let go

Frequently Asked Questions

Why does the GMAT® penalize you more for missing easy questions?

The GMAT® Focus Edition (as of 2026) uses an adaptive scoring algorithm. The difficulty of the questions you see is factored into your score, not just the number you get right. When you answer an easy question incorrectly, you signal to the algorithm that you can't reliably handle that difficulty level — and it adjusts accordingly. Missing hard questions has less impact because the algorithm expects some misses at the top of your range.

How do I convert decimals to fractions for GMAT® computation problems?

Move the decimal point right until you have a whole number, then put that number over the corresponding power of 10. So 0.045 becomes 45/1,000 (three decimal places = divided by 1,000). 1.9 becomes 19/10 (one decimal place = divided by 10). 0.03 becomes 3/100. Once everything is a fraction, flip the denominator fractions and multiply across.

What is the key insight on the 1/(2² × 3² × 4² × 5²) problem?

The key move is recognizing that 2² × 5² = 4 × 25 = 100. Since dividing by a power of 10 only shifts the decimal point — it doesn't introduce any new non-zero digits — you can set that factor aside. What remains is 1/(3² × 4²) = 1/144. Long divide 1 by 144. The result is 0.006944444..., which has three distinct non-zero digits: 6, 9, and 4.

How do you know when to let go of a hard GMAT® question?

Pay attention to the direction your confidence is moving. If you started a problem at 80% confidence and you've been working for 90 seconds without making clear progress — and your confidence has dropped rather than risen — that's a signal. Making an educated guess and moving on protects the time you need for problems where you do have traction. Missing a hard question has a smaller score impact than missing an easy one while rushing.

Should I use long multiplication or convert to fractions on GMAT® decimal problems?

Both work. Long multiplication is faster if you've built strong decimal intuition. Fraction conversion is more reliable if you haven't — it reduces most decimal problems to integer arithmetic and cancellation. If you're early in your prep and have time to build decimal skills, develop both. If you're close to test day and uncertain about decimals, fraction conversion is a defensible default for almost every computation problem you'll see.