What these guides are for
A score on its own does not tell you much. Two hundred and fifty milliseconds sounds fast until you learn that a large part of it is your monitor and your mouse rather than your nervous system, and eight clicks per second sounds impressive until you see how differently a one second burst and a ten second run behave. These guides give the numbers context, so that a result here means something beyond a position on a bar.
How the tests relate to each other
The benchmarks fall into three groups. Speed tests such as reaction time, click speed and the spacebar test measure how quickly you can produce a response, and their scores move with your hardware as much as with your reflexes. Memory tests such as the chimp test, number memory and sequence memory measure how much you can hold and for how long, and those scores are far more stable across devices. Attention tests such as dot count and the drift test sit between the two, measuring how much you can follow at once.
Comparing across groups is rarely meaningful. A strong reaction time says nothing about your digit span, and the percentile shown after a run is calculated against that test alone rather than against any combined ability.
Where the numbers come from
Each percentile is drawn from published distributions for the underlying task where one exists, and from the observed spread of ordinary play where it does not. They estimate where a score sits among people who take these tests. They are not clinical norms, none of this is a diagnostic instrument, and nothing here measures intelligence.
Your results stay in your own browser. Nothing is uploaded, and clearing your browser data clears your history with it.