Num Drill

Why Num Drill works the way it does

The short version: most of what makes Num Drill effective comes from well-studied learning science, not from anything we invented. This page explains the research behind the design choices, and what we’re explicitly not claiming.

The goal is fluent recall, not mindless clicking

Math fact fluency means a child can retrieve answers to basic arithmetic questions — 7 × 8, 13 + 9, 56 ÷ 7 — quickly and accurately, without counting on fingers or pausing to derive them. The 2008 U.S. National Mathematics Advisory Panel report identified fluency with whole numbers and fractions as one of the foundational skills that predict success in algebra and beyond.7

The point is not speed for its own sake. It is that fluent recall frees up working memory to handle harder problems. A child who has to derive 7 × 8 in the middle of a multi-step word problem is using cognitive bandwidth that a fluent peer is spending on the problem itself.

Retrieval practice beats re-reading

One of the most replicated findings in cognitive science is that actively retrieving information — pulling it out of your head, often described as “the testing effect” — produces stronger long-term memory than passively reviewing the same material. Roediger & Karpicke (2006) is a foundational paper on this for non-math content;4 the principle generalizes well to arithmetic facts. Num Drill is structured as a sequence of retrieval attempts, not a sequence of explanations to read.

Short, repeated sessions work better than cramming

Spaced (or distributed) practice — the same total amount of practice spread over multiple short sessions instead of one long one — consistently outperforms cramming in math-specific research. Rohrer & Taylor (2006) showed measurably better retention of mathematics knowledge with distributed practice relative to massed practice.3 The simplest practical takeaway: 10 minutes a day, most days, builds more durable fluency than one 70-minute Saturday session.

Fact fluency reduces working-memory strain

Ashcraft & Krause (2007) showed that working-memory capacity and arithmetic performance are tightly linked, and that math anxiety further consumes working-memory resources during computation.1 Geary (2011) followed children over five years and found that foundational numerical skills predict math achievement growth into later grades.2 The implication for an app like Num Drill: getting facts automated isn’t just convenient — it directly lowers the cost of the harder math built on top of them.

The level should feel challenging but doable

Robert Bjork’s work on “desirable difficulties”5 argues that learning improves when practice is hard enough to require effort but not so hard it produces failure cascades. Adaptive systems — ones that adjust difficulty based on recent performance — aim to keep practice in that zone. Num Drill’s weighted sampling does this for individual facts (the ones you keep missing surface more often) and the six-level grid does it for difficulty bands.

We deliberately avoid going further into deliberate-practice territory than the research supports. Ericsson, Krampe, & Tesch-Römer’s framework6 is foundational, but it was developed studying expert adult performers. Translating “deliberate practice” directly to elementary-school math is an over-extrapolation that researchers in the area generally caution against.

Why Num Drill uses short, focused drills

Fast answer-entry loop

Type the answer, hit Enter, next question. On a tablet, tap the numpad and submit. We don’t animate between problems, don’t show explanations after each answer, and don’t reward correct answers with cartoons. Friction reduces the rate of retrieval attempts; we’ve removed it where we can.

Difficulty progression

Six visible levels per operation, pitched 2–3 grades ahead of typical pacing. Mastery is shown explicitly: bronze for attempted, silver for sustained >70% accuracy, gold for >90% accuracy under target time. The level grid is the parent-facing progress map.

No unnecessary game layer

Coins, points, badges, characters, and streak-loss anxiety are all design choices that pull engagement away from the actual practice. We’ve seen no compelling evidence that they improve fact fluency on net. The product reward is a faster personal best and a visible mastery dot.

What we do not claim

Num Drill is a practice tool. It is not a substitute for math instruction, a curriculum, or assessment. We do not claim that using it produces specific test-score gains, grade-level advancement, or any other measurable educational outcome — because doing so would require controlled research we have not conducted. We will publish observational product data (e.g. “players who completed N quizzes saw their average accuracy rise from X% to Y%”) only with the clear caveat that it is observational, not causal.

If you’re running a real math intervention with a class or a co-op and want to evaluate whether Num Drill helps, the right study design is a brief pre/post fluency measure with a few weeks of consistent practice, parent consent, and plain-English protocol. We’re happy to discuss.

Sources

  1. Ashcraft, M. H., & Krause, J. A. (2007). Working memory, math performance, and math anxiety. Psychonomic Bulletin & Review, 14(2), 243–248. doi:10.3758/BF03194059
  2. Geary, D. C. (2011). Cognitive predictors of achievement growth in mathematics: A 5-year longitudinal study. Developmental Psychology, 47(6), 1539–1552. doi:10.1037/a0025510
  3. Rohrer, D., & Taylor, K. (2006). The effects of overlearning and distributed practise on the retention of mathematics knowledge. Applied Cognitive Psychology, 20(9), 1209–1224. doi:10.1002/acp.1266
  4. Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. doi:10.1111/j.1467-9280.2006.01693.x
  5. Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe & A. Shimamura (Eds.), Metacognition: Knowing about Knowing (pp. 185–205). MIT Press.
  6. Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363–406. doi:10.1037/0033-295X.100.3.363
  7. National Mathematics Advisory Panel. (2008). Foundations for Success: The Final Report of the National Mathematics Advisory Panel. U.S. Department of Education. Full report PDF
  8. Siegler, R. S., Duncan, G. J., Davis-Kean, P. E., Duckworth, K., Claessens, A., Engel, M., et al. (2012). Early predictors of high school mathematics achievement. Psychological Science, 23(7), 691–697. doi:10.1177/0956797612440101

Want to put this into practice? Start with multiplication, read about fractions practice for 4th graders, or jump to building a 10-minute routine at home.