Spaced, untimed fluency practice — free
Our practice portal lets students revisit foundational facts a little at a time, from memory, with instant feedback and no clock — exactly the recipe the research supports.
Open the practice portal →A great many students who "can't do" GCSE, IB or SAT maths can, in fact, do it — once you remove the slow, error-prone tax that shaky basics levy on every single line of working. This is the research on fluency: why small foundational gaps cascade so far, what fluency actually means, and how to build it without the timed-test fear that so often comes attached.
A sixteen-year-old comes to me "failing GCSE maths". We start a past paper. The algebra reasoning is fine — better than fine. But every few lines, the work slows to a crawl: a long pause over 7 × 8, a fraction added wrongly, a sign lost in the rush. By the end of the paper those tiny taxes have cost a dozen marks and most of the student's confidence. The diagnosis isn't "can't do Year 10 maths". It's a Year 4 foundation that was never made solid — and is now quietly sabotaging everything built on top of it.
This is one of the most common patterns in secondary maths, and one of the most overlooked, because the symptom (a poor GCSE mock) appears years and floors away from the cause (insecure basics). Fixing the visible topic rarely helps. Fixing the foundation often transforms the lot. Here is why the basics matter so much more than their humble status suggests — and how to rebuild them in a way that helps rather than frightens.
The tell-tale signs are specific. A student who reaches for a calculator for 6 × 7. Fractions that go wrong inside otherwise-correct algebra. Simple arithmetic slips that multiply across a long question. Negative-number errors that recur no matter how many times the topic is "covered". Individually each looks trivial. Collectively they decide grades — and, worse, they convince a capable teenager that they are "just bad at maths" when the truth is far more fixable.
Why does a small foundational gap do such outsized damage? Because mathematics is the most vertically stacked subject in the curriculum. Place value holds up fractions; fractions hold up algebra; algebra holds up calculus. A weak floor doesn't just wobble — it makes every floor above it wobble too, and the higher you build, the more the wobble shows.
Imagine reading a novel but having to stop and sound out one word in every sentence. You can read — but it's exhausting, slow, and by the end of the page you've lost the plot. A student who must recompute every number fact is reading maths one painful word at a time. They never get to enjoy or even reach the story, because all their effort is spent decoding what should be automatic.
That analogy is also the mechanism, and it links straight back to working memory. When a fact like 7 × 8 = 56 is instant, it occupies almost no mental space, leaving the workspace free for reasoning. When it must be reconstructed each time, it floods the workspace before the real problem even starts. Fluency, in other words, isn't about being fast for its own sake. It's about freeing the mind to think.
It's worth being precise, because the word causes arguments. The National Council of Teachers of Mathematics defines procedural fluency as the ability to apply procedures "efficiently, flexibly, and accurately" — to choose a sensible method, adapt it, and know when each one fits. Crucially, NCTM insists fluency should be built on conceptual understanding, not bolted on instead of it.
So fluency has three parts, not one. Accuracy (getting it right), efficiency (without laborious effort), and flexibility (choosing a smart route — knowing that 8 × 25 is easier as 2 × 100). Fast recall of facts supports all three, but "fast" alone was never the goal. A child who blurts answers quickly but can't choose a good strategy is not fluent; a child who reasons flexibly but recomputes every fact is not fluent either. We want both.
Here the evidence is unusually clear and unusually practical. Two of the best-established findings in the science of learning are tailor-made for number facts.
Rohrer and Taylor (2007) had students practise maths in different patterns. Practice spread out over time produced far better long-term performance than the same practice massed into one block — and mixing problem types ("interleaving") beat doing them in neat same-type sets. The counterintuitive part: spaced and mixed practice often feels harder and looks less impressive in the moment, yet it's what actually sticks. Most textbooks, which group practice by topic, accidentally optimise for the wrong thing.
Roediger and Karpicke (2006) demonstrated the "testing effect": the act of recalling something from memory strengthens it far more than studying it again. Pulling 7 × 8 out of your head — even effortfully, even with the odd error — builds the memory more than looking at a times-table chart ever will. For number facts, this means low-stakes self-quizzing is not just assessment; it is the most efficient form of practice there is.
Put together, these give a simple, evidence-based recipe: short, frequent, mixed, from-memory practice. A few minutes most days, recalling facts rather than re-reading them, with different facts jumbled together. It is almost the opposite of the once-a-week, one-topic, look-and-copy worksheet — and far more effective.
| Instead of… | Do this | Why it works |
|---|---|---|
| One long cram before a test | A few minutes most days | Spacing builds durable memory; cramming fades fast. |
| Re-reading the times-table chart | Recalling facts from memory | Retrieval strengthens memory far more than review. |
| One fact family at a time, in order | Mixed facts, jumbled | Interleaving forces the choice of method, like real problems do. |
| High-pressure timed tests | Low-stakes, untimed self-quizzing | Keeps the retrieval benefit without seeding anxiety. |
It would be dishonest to present fluency as a settled, friendly consensus. There is a real and useful disagreement about how to build it, and parents deserve to see both sides.
On one side, cognitive scientists and bodies like NCTM emphasise automaticity: instant recall of facts frees working memory, and retrieval practice is how you get there. On the other, educators such as Stanford's Jo Boaler argue that the way fluency is often pursued — especially timed tests — does real harm. Boaler's widely-read work contends that the onset of timed testing is, for a substantial share of children, the very beginning of maths anxiety, and that under time pressure working memory becomes blocked so that students cannot even access facts they know.
Notice that both sides actually agree on the core science: facts matter, and anxiety blocks working memory. The dispute is about method, not goal. You can hold both truths at once — fluency is valuable and timed tests are a poor, often harmful way to build it. That's exactly why the recipe above favours frequent, low-stakes, untimed retrieval: it banks the automaticity benefit the cognitive scientists want while sidestepping the anxiety the critics rightly warn about.
Foundational gaps are easy to acquire and easy to miss. A child off sick during the week fractions were introduced. A topic "covered" but never made automatic before the class moved on (the pace problem in action). A reliance on finger-counting or the calculator that papers over the gap just well enough to pass the next low-stakes check. Each leaves the floor a little weaker — and because the child can still limp through easy work, nobody notices until the load above gets heavy and the whole structure starts to lean.
Singapore builds fluency on a bed of number sense first — "number bonds" and mental strategies before drilling facts — so that recall and understanding grow together. Fluency is treated as the natural product of secure concepts, not a separate race, which is part of why Singapore tops international tables without relying on rote alone.
Japanese primary maths pairs strong fact fluency with reasoning, often through a single rich problem explored deeply. Mental calculation is valued, but it sits inside lessons about why methods work, so fluency and understanding reinforce each other rather than competing.
England introduced a national Multiplication Tables Check for Year 4 — an explicit bet that times-table fluency by age nine is worth measuring. Supporters cite the working-memory payoff; critics raise exactly the timed-test anxiety concerns above. It's the live disagreement, made into national policy.
The US is where the fluency argument is fiercest — the "math wars" between fact-drilling and reform approaches. Boaler's anti-timed-test work is influential there, as is the counter-argument from cognitive scientists. The emerging middle ground is retrieval practice without the stopwatch.
India has a strong cultural tradition of memorising tables, often well beyond 12 — producing impressive recall, but sometimes, critics note, ahead of the understanding that should underpin it. It's a vivid case of fluency and conceptual understanding needing to be developed together, not one at the expense of the other.
Design practice for durability, not for looking good today. Spaced, interleaved, retrieval-based practice often produces messier in-lesson performance and better long-term retention. Build in low-stakes recall starters that revisit old facts, and resist the tidy single-topic worksheet that feels productive but fades fast.
Separate fluency-building from assessment. Use frequent, no-grade quizzing as practice, not judgement. This captures the testing effect while removing the threat — the practical resolution of the Boaler/NCTM debate.
Diagnose the foundation before re-teaching the topic. When an older student fails a topic, check the basics underneath it first. Patching a Year 4 gap frequently does more for a Year 10 grade than re-teaching the Year 10 content ever could.
Our practice portal lets students revisit foundational facts a little at a time, from memory, with instant feedback and no clock — exactly the recipe the research supports.
Open the practice portal →"Calculators mean kids don't need number facts any more."
Facts matter for working memory, not arithmetic. A child recomputing basics has no spare capacity for the reasoning that calculators can't do for them.
"Fluency just means being fast."
Fluency is accuracy, efficiency and flexible strategy choice, built on understanding. Speed alone, without sense, is brittle.
"Timed tests are the way to build times-table fluency."
Timed tests can trigger anxiety that blocks the very facts being tested. Frequent, low-stakes, untimed retrieval gives the benefit without the fear.
Yes — but for a subtler reason than arithmetic. When facts are instant, they take up almost no room in working memory, leaving space for the reasoning harder maths demands. A child who recomputes 7 × 8 mid-problem is spending mental capacity they badly need elsewhere. The calculator handles the sums; it can't handle the thinking.
Not just speed. NCTM defines it as applying procedures efficiently, flexibly and accurately, and knowing when to use which. Fast recall is one ingredient, but choosing smart strategies and understanding why they work matter just as much. A fluent student is quick and sensible and right.
Genuinely debated. Cognitive scientists value automaticity and retrieval; critics like Jo Boaler argue timed tests seed anxiety and don't measure real fluency. The sensible middle path — and the one we use — is frequent, low-stakes, untimed retrieval: you keep the memory benefit and lose the fear.
Short, frequent sessions from memory, with facts mixed together. Spacing (spread over days) beats cramming, retrieval (recalling) beats re-reading, and interleaving (jumbling facts) beats neat single-table sets. A few calm minutes most days is the whole recipe.
Very often, yes. Advanced topics assume instant access to fractions, tables and number sense. When those are shaky, every higher problem slows down and accumulates errors, so the marks lost look like topic failures when the real cause sits years below. Diagnosing and quietly patching the foundation frequently unlocks the rest.
The free assessment includes a quick diagnostic of the basics underneath the visible topic — so you find out whether the real fix is years below where the marks are being lost.
Book the free assessment