N-Back – A Minimal, Adaptive Dual N-Back Game for Brain Training

Hi HN,

I built N-Back.net, a lightweight and responsive Dual N-Back task for working memory training.

I was frustrated by the clunky UI and heavy design of most existing n-back tools, so I made something clean, fast, and distraction-free. No login, no ads, no tracking – just cognitive training in your browser.

Key features: • Dual mode: visual + auditory • Adaptive difficulty: level up as you improve, drop down if needed • Dark mode and keyboard-only mode • Clearly differentiated letter sounds, carefully selected to avoid confusion • Performance tracker (WIP: I’m working on streaks and detailed history) • Mobile-friendly (works well on phones & tablets)

I’m planning to add: • Custom training modes (single-modality, color, etc.) • User-defined N range • Optional accounts for progress saving (but always optional)

Would love your feedback on: • UX flow and performance • Cognitive challenge balance • What would make you come back and use it regularly?

Site: https://n-back.net

Thanks for checking it out!

My personal experience with n-back mainly revolves around discovering the large variation in my ability to focus. Practisibg does get me better at the n-back task (not anything else as far as I can tell) but daily form determines results to a large degree. This effect seems to transfer somewhat: on weeks I'm overall worse at focusing on n-back I appear to also produce fewer commits.

But what determines this daily form I don't know. I suspect sleep, stress, and exercise has something to do with it.

I've been using the n-back protocol for over a year and it's definitely improved my cognitive abilities. I don't know if my IQ has increased, but on days where I use this protocol(20-25 minutes) I see a marked improvement in the clarity and vibrancy of my thoughts- put another way this protocol's biggest benefit is it's effectiveness in clearing the mental fog in my head. I've tried experimenting with other protocols(solving crossword/chess puzzles, meditation,etc..)but they were not nearly as effective as n-back.

I've actually built my own variation of n-back(triple n-back), where you're tested on the colour of the stimulus in addition to the sound and position-so it's much more cognitively demanding. If you're interested, you can check it out here: https://mind-workout.pages.dev

Generalization beyond n-back tasks had not been established, at least at the last time I looked into the field, maybe 8 years ago. There was always the hope that tasks that focused on specific core cognitive processes would lead to broad cognitive improvements, but I am not too confident about it.

One problem in this field has been weak control conditions..e.g. no train conditions. I often thought that a cardio control conditions would be useful. I'd almost guarantee that for most people, 60 minutes a week jogging would lead to better cognitive improvements than 60 minutes of n-back.

This is very nice! But I don’t quite get how 70% is considered good performance — I managed to outperform that by literally doing nothing:

Game Complete!

Accuracy: 78.6% Correct Responses: 11 / 14 Average Reaction Time: 0 ms

I also wrote a minimal version for me, https://mostlymaths.net/nb/

Only works well on mobile portrait though, I designed it especifically for my phone. Don't use it much though.

" Works on All Devices" statement doesn't seem to be true - I cannot press "space key" on my iPad (Safari).

It’s hard to tell when a turn starts if the tile stays on the same square. Could you possibly add a quick fade animation to the tile?

"Working Memory Test" link returns 404

Ahh, "brain training" - games you play instead of getting better at value added tasks. This shit doesn't generalize and the opportunity cost results in making you dumber, not smarter.

[dead]