![Image 1 — [Android Beta] I built a free, offline cognitive trainer. Need DNB veterans to test my adaptive timing (is 1.7s too fast?)](https://preview.redd.it/pzbekn7xm19h1.png?width=1080&format=png&auto=webp&s=cddd1f1de76b45a99a09dc5c7b66eb33b10c15e0)
![Image 2 — [Android Beta] I built a free, offline cognitive trainer. Need DNB veterans to test my adaptive timing (is 1.7s too fast?)](https://preview.redd.it/b5jfw4bxm19h1.png?width=1080&format=png&auto=webp&s=3383b8d08a11681c48fc11f8b94b29c36924168f)
[Android Beta] I built a free, offline cognitive trainer. Need DNB veterans to test my adaptive timing (is 1.7s too fast?)
Hi everyone,
I’ve been working on Kognix, a neurocognitive trainer focused on working memory and executive functions. It’s 100% offline, completely free, with no ads and zero telemetry.
The core modules (Adaptive Dual N-Back, Stroop, Digit Span, eCorsi, Memory Palace) are functional, but I’ve hit a wall with calibration and I need objective data from people who actually play these tests regularly.
Specifically, I need your feedback on the Dual N-Back Adaptive Algorithm:
Currently, the adaptive difficulty drops the timing to 1.7s per turn at higher levels. Some initial feedback suggests this is too aggressive for a mobile UI and turns it into a reaction-time test rather than working memory. I would love to know how this pacing feels to experienced DNB players here. Is 1.7s playable on a touch screen, or should I hard-cap it at 2.0s?
(Secondary ask: The app generates a KCI radar chart based on your baseline. I need fresh eyes to tell me if the psychometric normalization feels balanced).
How to test (Android only):
You can download the APK directly from my GitHub Releases page here:
GitHub Release v0.9.0-beta (Download APK)
(Note: I have temporarily enabled Wiredash in this beta so you can send me feedback directly inside the app without leaving it. The final release will be entirely offline).
Any harsh critiques on the UI or the timing limits are highly appreciated. Thanks for your time!