How the games and tests work

NeonReflex is a set of quick games and cognitive tests you can play in a browser tab. Some are pure fun, some measure something real about reaction speed or memory, and a few are both. This page explains how they are built and scored, because a benchmark you cannot see the rules of is not a benchmark, it is a slot machine.

The rules are written down, in full

Every game page states its actual mechanics: how many lives you get, exactly how a pattern grows from one level to the next, what counts as a mistake, when the run ends, and how the final score is calculated. These are not vague descriptions written to sound plausible. They are drawn from the same code that runs the game, so when a page says the visual memory test adds the level number plus two tiles each round, that is the literal rule the game follows. If you want to know why you lost, the page can tell you.

Score interpretation is honest about what it can claim

Where a page suggests what a good score looks like, it is describing rough, typical ranges, not a clinical norm. A reaction-time result depends on your hardware, your display's refresh rate, and how awake you are, and the page says so rather than dressing the number up as a diagnosis. The cognitive tests here measure a narrow, specific thing, recognition memory or visual span or reaction latency, and each page is careful to say what that thing is and, just as importantly, what it is not. None of these games diagnose anything. They are practice and curiosity, not medicine.

Your scores stay on your device

Best scores and streaks are saved in your own browser using local storage. They are not uploaded, not attached to an account, and not compared against strangers on a server, because there is no server keeping them. That is a deliberate choice. It means your record follows the browser rather than you, so clearing site data resets it, and it means there is no global leaderboard to game. The number you beat is your own previous number, which is the only honest opponent a solo benchmark has.

No dark patterns

There are no fake "you almost had a high score" nudges, no artificially inflated difficulty to sell you something, no countdown timers manufactured to create urgency. The randomness in these games is real randomness, which is why a lucky draw genuinely is easier and an unlucky one genuinely is harder, and the pages say when that variance is part of the design. The site is free, supported by a few display ads placed away from the game area so they do not interfere with play.