HumanSpec Science
Simple tests. Clearer context.
Human performance is affected by more than ability. We design short browser tasks that make the interaction easy to understand, the score easy to repeat, and the limitations hard to miss.
What you can measure
Four focused signals, one honest frame.
These tests are intentionally narrow. That makes them easier to play and gives you a better chance of understanding what changed between sessions.
Speed and response
Reaction time and click speed capture how quickly you respond to a visible cue and coordinate an input.
Working memory
Number memory asks you to hold a short sequence briefly and reproduce it after the visual cue disappears.
Visual recall
Visual memory combines attention, spatial encoding, and the ability to rebuild a pattern after a short delay.
Typing control
Typing speed pairs a pace measure with accuracy so a fast result does not hide avoidable errors.
Our design principles
Useful beats impressive.
A precise-looking score can still be misleading if the task hides its assumptions. We keep the experience lightweight and the interpretation grounded.
One task at a time
Each test focuses on a small, understandable interaction instead of pretending to measure the whole mind.
Show the rules
The task, scoring signal, and common sources of variation are explained on the same page as the game.
Prefer a range
Repeated attempts under consistent conditions are more useful than a single impressive outlier.
How to read a result
A result is a snapshot of one task, on one device, at one moment. Use it to build a personal range, notice a change, or guide practice.
For a fair comparison, keep the browser, display, input device, and environment consistent. Run several attempts and pay attention to the middle of the range.
What these tests cannot tell you
HumanSpec does not diagnose a medical condition, measure intelligence, or predict job, academic, or athletic performance.
Browser timing also includes device and network behavior. We describe those limits because transparency is part of a useful test.
Editorial method
Evidence, product data, and limits stay separate
Research sources
Articles use primary papers or established technical references for task concepts. A source supports a claim; it does not validate every online implementation.
Product references
HumanSpec ranges describe this site’s browser tasks. They are not a representative sample or a population benchmark.
Review standard
The HumanSpec Editorial Team checks that instructions, scoring language, links, and limitations match the experience users can actually run.
Starting sources
Continue reading
Notes from the HumanSpec journal
Performance
How to measure reaction time consistently
Why a repeatable personal range is more useful than one unusually fast score.
Measurement
Browser timing and display latency
How device and browser conditions affect fair comparisons.
Baseline
Build a personal performance baseline
Track trends across repeated sessions without over-interpreting noise.