Skip to content

Saying It Backwards: Why Reversal Costs You Digits

Reversing a sequence adds a second job on top of holding it. Why backward span runs shorter than forward span, and what that gap is actually telling you.

By Published Updated 5 min read

Reign Creative LLC makes IQ Test & Brain Training Games, which this article links to, so we have a commercial interest in it. How we write and check articles.

Key takeaways

  • Forward recall asks you to hold a sequence. Backward recall asks you to hold it and transform it, which is a second job running on the same limited resources.
  • Türüdü and colleagues describe the forward task as primarily measuring the storage capacity of the phonological loop, while the backward task engages the central executive for active manipulation of information.
  • The standard clinical procedure recognises the difficulty gap in its own design: Choi and colleagues describe forward recall beginning at a length of three and backward recall beginning at a length of two.
  • A random-direction version, where you only learn which way to answer after the digits disappear, is harder than either fixed direction because it blocks you from choosing an encoding strategy in advance.
  • IQ Test & Brain Training Games keeps your best forward span and your best backward span as separate personal bests. Those are game statistics, and the app's memory test and IQ-style test produce results that its listing calls estimates based on app activities, not clinical or professional assessments.

Reversing a sequence is harder than repeating it because you are doing two jobs with one set of resources: holding the digits, and rearranging them before you answer. Türüdü and colleagues put the distinction plainly in a 2026 study in iScience — the forward task "primarily measures the storage capacity of the phonological loop", while the backward task "requires individuals to recall the sequence in reverse order, which engages the central executive for active manipulation of information." The gap between your two spans is not a defect. It is the task working as designed.

The design already assumes the gap

You can see the assumption baked into the standard clinical procedure. In their normative study, Choi and colleagues describe presenting digits at one per second, beginning at a length of three for forward recall and a length of two for backward recall, with two trials at each length and a halt when both trials at a length fail.

That one-digit head start is an admission. The people who built the instrument did not expect the two directions to be equivalent, so they did not start them in the same place. If you have been quietly disappointed that your backward runs come up short, the answer is that everyone's do.

What nobody can honestly give you is a fixed size for the difference. It moves with age, with strategy, with how long the sequence is, and — as the same iScience work found — with how the task is delivered. A gamified online version produced significantly higher scores than both onsite and standard online versions, and crucially the boost was identical for forward and backward tasks. A friendlier interface lifted both numbers without touching whatever sits underneath.

Storage and manipulation are different demands

The useful way to think about this is not "backward is harder" but "backward adds a transformation". Holding 4 9 1 7 is a storage problem. Producing 7 1 9 4 requires you to keep the original available while you walk it in the other direction, and the walking itself consumes the resource the holding depends on.

Nelson Cowan's review of memory terminology is helpful here, because it explains why the field keeps circling this distinction. He describes working memory as having been "conceived and defined in three different, slightly discrepant ways: as short-term memory applied to cognitive tasks, as a multi-component system that holds and manipulates information in short-term memory, and as the use of attention to manage short-term memory." Backward recall sits squarely in the second and third of those. Forward recall does not have to.

We treat that contrast at length in our article separating short-term memory from working memory, and the capacity question itself is handled in the working memory guide, which is the anchor for everything in this hub.

Strategies that move a backward span

Because the difficulty is transformation rather than storage alone, the strategies that help are the ones that reduce how much you have to transform.

Reverse groups, not digits. Take 6 2 8 3 5 1. Held as six items and reversed one at a time, you will lose the middle. Held as 62 83 51, it is three units. Reverse the units to 51 83 62, then reverse inside each to get 15 38 26. Read out, that is 1 5 3 8 2 6 — which is the original read right to left. You performed two small reversals instead of one six-step one.

Encode positions, not just values. People who do well often report anchoring the first and last items immediately, because those are the two that the reversal makes most important. The interior can then be rebuilt around fixed endpoints.

Say it once, then stop saying it. Rehearsing forward while trying to output backward is the most common self-inflicted failure. One clean pass to encode, then switch modes.

These are genuine techniques, and they will improve your score. They are also the reason to be careful about what the improved score means. Norris, Hall and Gathercole, testing extensive digit span training directly, concluded that such training "does not substantially improve the capacity of verbal STM", with little evidence of fundamental capacity change and minimal transfer to related tasks such as letter span. Better strategy on a practised task is a real and useful thing. It is not the same as a bigger store, and we do not claim it is.

The version that removes your strategy

There is a third format worth knowing about, and IQ Test & Brain Training Games — our own app, built by Reign Creative — ships it: a random-direction drill where you see the digits and only find out afterwards whether to type them forward or backward.

This is harder than either fixed direction, and the reason is instructive. In a fixed backward task you can choose an encoding that suits reversal from the first digit. In a random-direction task you cannot commit, so you either encode twice or encode neutrally and pay the transformation cost at the end. It is a neat demonstration that a large part of span performance is preparation rather than raw holding.

Alongside it the app ships a plain reverse-sequence mode and a backward-span ladder that mirrors the forward one, and it stores your best forward span and your best backward span separately rather than collapsing them into a single figure. That separation is the right design for practice, because the two numbers move for different reasons.

What these numbers are, and what they are not

They are game statistics. They were produced on your own device, at a time you chose, with no examiner, no controlled presentation and no reference population. They are good for one thing: comparing this week's runs to last week's on the same mode at the same difficulty level.

They are not a measurement of your working memory, not a cognitive assessment, and not comparable to anything a qualified professional would produce with a standardised instrument. If a real question about memory or attention is troubling you, that is a conversation for a professional, not for a phone game. The same limits apply to the app's memory test and to its IQ-style test, which covers working memory among other areas: the listing calls their results estimates based on app activities, not clinical or professional assessments.

Used within those limits, backward drills are an unusually clear window into your own processing: you can watch the point at which adding a transformation breaks a sequence you could otherwise hold. The app's memory games and tests are described on the IQ Test & Brain Training Games app page, the rest of this subject in the memory and attention hub, and the wider set of learning articles in the education and brain section. If you have not yet read how a span figure is produced in the first place, start with what a digit span number actually means.

Good to know

Frequently asked questions

Why is backward digit span harder than forward?
Because it asks for two things at once. You still have to hold the sequence, and now you also have to reverse it before you speak or type it. Türüdü and colleagues describe the backward task as engaging the central executive for active manipulation of information, where the forward task primarily measures the storage capacity of the phonological loop. The transformation competes with the holding.
How much shorter is a backward span than a forward span?
There is no single number, and anyone quoting one is overreaching. The gap varies with age, strategy, sequence length and how the task is delivered. What is consistent is the direction of the difference and the fact that test designers assume it: the standard clinical procedure starts backward recall one digit shorter than forward recall.
What does backward span measure that forward span does not?
Manipulation rather than pure storage. The literature often treats forward recall as loading on a short-term storage factor and backward recall as loading alongside working memory measures that require processing as well as holding. That is a claim about what the task taxes, not a verdict on a person, and it is contested at the edges.
Can you get better at reversing a sequence?
You can get better at the task. The usual route is strategic: break the string into two or three groups, reverse the groups, then reverse within each group, so you are transforming three chunks rather than eight digits. Gains of this kind are specific to the practised task and reflect better technique rather than a larger underlying capacity.
Does the app test my working memory?
Not in the clinical sense. IQ Test & Brain Training Games includes a memory test and an IQ-style test whose areas include working memory, and its Google Play listing describes their results as estimates based on app activities, not medical, psychological, clinical or professional diagnostic assessments. Nothing about a phone session is standardised administration: you choose the time and the conditions, and nobody observes. A standardised psychological assessment is a different thing entirely, administered by a qualified professional under controlled conditions.

Sources

  1. Platform, modality, and gamification effects on digit span task performance (opens in a new tab)iScience, via PubMed Central (PMC13214330) — accessed
  2. A Normative Study of the Digit Span in an Educationally Diverse Elderly Population (opens in a new tab)Psychiatry Investigation, via PubMed Central (PMC3942550) — accessed
  3. What are the differences between long-term, short-term, and working memory? (opens in a new tab)Progress in Brain Research, via PubMed Central (PMC2657600) — accessed
  4. Can short-term memory be trained? (opens in a new tab)Memory & Cognition, via PubMed Central (PMC6647533) — accessed

About the author

— The independent app studio that builds the apps covered on this site and publishes every article here. On Google Play it publishes as Reign Collective Apps.

We use AI tools to help research, draft and edit our articles. Each article lists the sources its factual claims rely on, and details about our apps come from their live Google Play listings. No article on this site has been reviewed by an independent expert unless the reviewer is named on it.

Spotted an error? Tell us — how corrections are handled is set out in our editorial policy.

Keep reading