CompetitiveEasy5 min

Reflex Profile

Five reaction tests in five minutes — and a profile of how you respond.

By Dmitry Baluev · Updated May 16, 2026

About this test

Five short sub-tests measure speed, choice, sound, restraint, and switching. We turn the five scores into one Reflex profile — Hummingbird, Falcon, Metronome, Sentinel, or Sleeper. For best results, use a desktop with a wired mouse on a 120Hz+ display. Touchscreens, trackpads, and Bluetooth peripherals add latency the test can't subtract.

Start the battery
#reaction#reflex#battery#go-no-go#task-switching#5-tests

What your result could be

Every possible result for this test, and what each one means.

Hummingbird — Micro-twitch speed across the board.
You react before your conscious mind catches up. The green appears and your finger is already moving — there is no measurable gap. People who watch you play games or sports notice you reaching for things a beat earlier than seems possible. Your nervous system is wired for low-latency response: when something changes in your field of view, your motor cortex fires before your prefrontal cortex has time to deliberate. Sub-200 ms simple visual reaction is roughly the top 3% of adults, and the Hummingbird profile is about producing that kind of response consistently rather than as a one-off lucky trial. The archetype name comes from the overall level rather than any single scale — how your own five sub-tests landed is set out below. Where this shines is the narrow but vivid family of domains where milliseconds genuinely decide outcomes. Competitive video games at the professional aim-trainer level, sports requiring split-second timing (boxing, table tennis, fencing, the goalkeeper's dive, the wide receiver's burst), live music improvisation where the next note has to be inside the half-beat, emergency-response work where the speed of the first move matters more than the deliberation behind it. It's also a real component of safe driving in dense urban traffic — you literally see and respond to brake lights faster than other drivers. The blind spot is the inhibition cost. The Hummingbird profile is built on the pure-speed sub-tests (Lightning, Fork), and what such wiring pays most for is the inhibition tests (Stop-Signal) where the right answer is 'don't act' — how those sub-tests landed for you specifically is set out below. In daily life it looks like this: you sometimes act on stimuli that turned out not to require action — false-start, move first and verify second, double-spend on cleanup of the wrong commitment. The same wiring that gives you the speed advantage makes it harder to hold when the situation rewards waiting. The growth line is inhibition — building the muscle to hold when the situation calls for holding. The raw micro-twitch is a real and rare capability; the discipline that pairs with it is what turns it into a long-term advantage rather than a chronic source of small avoidable errors.
Falcon — Focused predator — fast and accurate, locked-in.
You combine speed with discrimination: you react fast AND react correctly, and the score is earned in the decision component of reaction, not just the raw motor speed. Falcon is the name for that level, and what usually carries it is the pairing of discrimination and inhibition — not tapping on red when it's hard to hold, and not drifting by the fifth minute where many people drift after the second sub-test. People who work with you notice that you don't fumble decisions under time pressure — you commit cleanly and you commit correctly. The archetype name comes from the overall level rather than any single scale — how your own five sub-tests landed is set out below. Where this profile shines is in roles that require both speed and accuracy simultaneously under sustained attentional load. Surgical and technical work where the wrong move can't be undone, financial trading desks where you're choosing among several plausible actions in seconds, air traffic control, emergency-medicine triage, high-stakes negotiation where reading the room and responding correctly happens in the same beat, motorsport, professional esports at the team-strategy layer rather than the pure-aim layer. The Falcon advantage is not 'faster than everyone' — it's 'fast and correct when both have to be true and the cost of either failure is high.' The blind spot is attentional fatigue from sustained focus. The Falcon profile is the most prone of the five to burnout if you treat every situation as high-stakes locked-in time; the cognitive overhead of maintaining vigilance + discrimination at peak is expensive in a way that doesn't show up on a single five-minute test. The other failure mode is under-investing in the 'low-power coast' mode that humans actually need for recovery — defaulting to focus when the situation doesn't require it, then crashing on the situations that do. The growth line is selective intensity — learning to drop into the focused-predator mode when it matters and out of it when it doesn't. The capability is real and rare; protecting it from constant deployment is the meta-skill.
Metronome — Consistency over peak speed.
Your reaction times are not the fastest, but the composite is built on evenness rather than speed. Where Hummingbird and Falcon profiles pull ahead on spikes, Metronome gets there on a near-flat line of mid-range responses. This is its own kind of capability: in most real-world domains, reliable performance beats peak performance. The colleague you can count on for the 80% solution at the deadline is more valuable than the genius who delivers 110% half the time and 50% the other half. The Metronome profile reads as the adult-professional baseline — solid, repeatable, predictable. The archetype name comes from the overall level rather than any single scale — how your own five sub-tests landed is set out below. Where this shines is in any role rewarding reliability over heroics. Engineering, accounting, surgery, teaching, music-ensemble performance, long-distance driving, ambulance dispatch, machining, anything where the same task happens many times and the expected error rate matters more than the best-case speed. The Metronome profile lives in the sub-tests where consistency matters more than peak velocity — Zebra (Task Switching), where switch cost stays low because you stay in process when the rule changes rather than spiking and crashing; Echo (Audio), where attentional drift is the failure mode for others but the ear stays on the channel. Teams notice that you don't crash projects with hero-or-zero variance; the work just gets done at a steady cadence. The blind spot is invisibility in metric-driven environments that reward peaks. Annual-review systems built around 'breakthrough moments' or 'leaderboard top-N' often fail to credit the Metronome contribution that prevented twenty would-be breakdowns nobody saw. You may also under-market your own reliability because it doesn't feel like a story worth telling — and miss opportunities that genuinely require an asymmetric speed burst (a one-time creative bid, a chance to step into a faster lane). The growth line is storytelling about consistency — learning to articulate the value of reliability in a culture that often defaults to peak-worship. The work is the same; the framing changes who notices.
Sentinel — Vigilance over speed — never misses the stop signal.
Sentinel is the name for a specific cognitive style: verification-first action. Speed and accuracy are tied together here by one decision — to wait, letting a stimulus register fully before committing. In the green-screen test that costs 50–80 ms, because you wanted to be sure it really was green; in exchange, the finger doesn't go to red even after a long string of greens that have trained the reflex toward 'tap.' The style has a name in the academic literature: 'response cautiousness.' It's not slowness; it's a different setting on the speed-accuracy tradeoff dial. The archetype name comes from the overall level rather than any single scale — how your own five sub-tests landed is set out below. Where this profile is the natural fit is in roles where the cost of a false move is high and the cost of a slightly slow move is low. Safety-critical work — aviation, surgery, nuclear-plant operations, hostage negotiation. Quality assurance, security operations, auditing, proofreading, code review. The doctor who double-checks the chart, the auditor who spots the typo in the spreadsheet that everyone else's eyes slid past, the parent of a small child who never leaves the gate unlatched. Slower-but-correct dominates faster-but-wrong in any real domain where errors have consequences that don't go away by being fast. The blind spot is decision-paralysis in time-pressured environments where the cost of error is actually low — situations that genuinely reward 'act fast, fix afterward.' Sentinels can also over-check when the stakes don't warrant it, draining their own attention budget and frustrating faster collaborators. The other failure mode is mistaking caution for capability across the board: refusing first-mover opportunities in contexts where being first is more valuable than being right. The growth line is calibrated risk-acceptance — learning to distinguish situations where error has lasting consequences from situations where it doesn't, and choosing the appropriate speed-accuracy setting for each. The vigilance instinct is genuine strength; the meta-skill is knowing when to let it lead and when to override it.
Sleeper — Underrated player — deliberate, calm, hard to surprise.
Your composite is slower than average — and this is not the worst news. 'Sleeper' is the word for an underrated player in chess, in stock-picking, in basketball drafts: someone whose value isn't visible in the surface-level speed metric but who shows up reliably when it matters. Your nervous system is wired for deliberation over reflex. You're harder to surprise because you don't depend on twitch-speed to navigate the world; you depend on understanding what's coming. Where this asymmetric strength tends to land is in domains where the time-frame of relevance is hours to years, not milliseconds. Strategy work, long-form writing, design, research, code architecture, counseling, teaching, careful diagnostic professions — the situations that reward depth over instant motor response. People with this profile are often the ones who notice the second-order consequence three steps ahead while the rest of the room is reacting to the first thing on the agenda. Twitch-speed is a real capability for a narrow set of situations; analytical patience is a real capability for almost every situation that has consequences beyond the next ten seconds. The blind spot here is internalizing the speed metric as a verdict on personal worth — letting one millisecond number become the central self-story. It often shows up as quiet under-investment in the situations where speed actually would help (driving safely on the highway, the occasional fast game with friends) and over-investment in proving the slowness was 'really fine.' Both are tells that the metric matters more than it should. The growth line is separating capability from speed metric. The reaction-time score is one narrow measurement of one narrow capability — not a measure of intelligence, attention overall, or personal value. Treat this result as you would your height: a fact about you, useful when relevant, irrelevant most of the rest of the time.

Frequently asked

How accurate is this test?

Treat it as a quick snapshot, not a clinical measurement. Reflex Profile is five short tasks in a browser — useful for self-discovery and comparison, not for diagnosis. Your score in a single session reflects your device, your network, how rested you are, whether you've had caffeine, and dozens of other factors. For a more stable picture, run the battery two or three times across different days and look at the pattern, not the single number.

What is reaction time and why does it matter?

Reaction time is the gap between a stimulus appearing and you responding to it. The chronometry tradition starts with Donders in 1868 and runs through Hick (1952), Posner (1967), and Logan (1994). Reaction time is one narrow window into how the perceive–decide–act loop works under time pressure. It matters in driving, gaming, sports, and emergency response — but it is one component of cognition, not a measure of intelligence, attention overall, or general ability.

Why is my score much worse on mobile than on desktop?

Touchscreens add roughly 30–80 ms of latency vs a wired mouse; cellular networks add 20–100 ms vs ethernet; lower-refresh-rate displays (60 Hz vs 144 Hz) add up to 16 ms per frame. Combined, mobile users systematically score 50–200 ms slower on the exact same sub-tests. We show device class next to your score so you can compare honestly. Per-device leaderboard filtering is a v2 plan.

What if I scored about the same on all five sub-tests?

That's the Metronome profile — and it is a real result, not a missing one. Trial-to-trial consistency is a separate capability from peak speed, and in many real-world domains it matters more. A reliably good engineer, surgeon, or driver tends to be a Metronome. If your numbers are also high across the five, you may sit at the boundary between Metronome and Falcon.

Can my reflexes improve with practice?

Mildly. Direct practice on a specific reaction task yields about 5–10 ms of improvement that stabilizes after a few sessions and is largely task-specific. Sleep, hydration, caffeine, and being warmed up have a much larger same-day effect — easily 30–50 ms — and that is what most of the day-to-day variation in your scores is actually measuring. We do not claim that this test 'trains your brain' or improves real-world performance, and you should be skeptical of products that do.

Why don't you have a single global leaderboard?

Per-sub-test leaderboards exist and are filtered server-side against impossible-human submissions (faster than 100 ms repeatedly, automation patterns). We deliberately don't lead with a single cross-device global leaderboard because input-device variance — touchscreen vs mouse, 60 Hz vs 144 Hz display, wifi vs cellular — makes that comparison noisy enough to be misleading. Device class is shown next to every score.

How is this different from HumanBenchmark or BrainHQ?

HumanBenchmark is one single-test page per task (their Reaction Time is the famous one) — fast, free, but no composite across tasks and no profile, just a number. BrainHQ is a paid subscription brain-training product oriented around exercises and longitudinal scoring; not designed as a one-session profile. Reflex Profile sits in between: a free five-test battery in one sitting that returns one shareable archetype (Hummingbird, Falcon, Metronome, Sentinel, or Sleeper) plus per-sub-test percentile bars.

Is this an ADHD test or a concussion test?

No, and please do not use it as one. Reflex Profile is an entertainment self-discovery experience, not a medical device. It does not screen for ADHD, post-concussion syndrome, dementia, or any other condition. If you have concerns about your attention or reaction speed in daily life, please talk to a qualified clinical professional — a neurologist or neuropsychologist can run validated instruments that this test cannot replace.

Related tests

All Reaction & reflex tests