You get better at what
somebody is beating you at.

You type 74 words a minute. Is that good? You genuinely cannot tell, and neither can the tool that told you, because a score means nothing without a distribution behind it. Most practice sites stop exactly here: they measure you accurately and then leave you holding a number you have no way to read.

Every finished run here is told where it sits among all the runs on its board, in the same round trip that saves it. That is the difference between 74 wpm and 74 wpm, which beats 61% of runs and is 38 off the record. The second one tells you what to do next.

Deliberate practice needs two things that are hard to manufacture: immediate feedback, and a target just above your current level. A person who is a little better than you supplies both, for free, without anyone designing a syllabus. Losing narrowly is the most informative result there is, because it is the only one that proves the gap is crossable.

So the unit here is not a lesson, it is a link. Post a score, send it, and whoever opens it plays the identical test. In typing they race your actual keystrokes, replayed at the speed you typed them, which turns an abstract number into something you can watch pull away from you.

This is the part most competitive products get wrong, and it is why we refuse several things that would obviously increase engagement. There is no ELO here. No rank that can fall. No matchmaking, no seasons, no badge you lose by not showing up.

Ranked ladders work by making the number hurt, and a number that can hurt is a number people eventually stop looking at. Every comparison on this site is opt in: you choose to send a link, you choose to open the leaderboard. The daily resets at midnight whether it went well or badly, which is the point of it.

A five second click run and a sixty second one are not the same skill. Audio reaction beats visual reaction by physiology rather than by practice. So each of those keeps its own board, and there is deliberately no combined ranking anywhere on the site: adding 8 clicks per second to a 240 ms reaction would produce a number nobody earned.

The same rule is why the score is recomputed on the server from your raw input log rather than taken from the browser. A leaderboard is only worth looking at if the numbers on it are real, and a number you can type into a console is not a number.

Every mechanic here is one somebody actually relies on, in a game or at a desk, and each test says which:

• Typing · Anyone who works at a keyboard. Code, email, chat, notes: it is the most used mechanic at a desk, and the one people practise least.
• CPS · Games that ask for fast repeated inputs, and a plain honest measure of hand speed.
• Reaction · Anything where the cue is sudden: shooters, driving, sport, and every game that rewards seeing it first.
• Aim · Shooters above all, and cursor control generally. Precise fast pointing is as much a desk skill as a gaming one.
• Keyboard · Checking your hardware before you blame your hands. A dropped key costs a fight and a sentence alike.
• Trading · Holding a decision while the number moves. The discipline half of trading, practised without any money on it.

Typing is the clearest case. It is the single most used mechanic in office work and the one almost nobody deliberately practises, which is why the gap between a slow typist and a quick one is measured in hours per week.

Getting quicker at clicking will not make you good at a game, and we are not going to pretend otherwise. Mechanics are the floor, not the ceiling: they decide what your decisions are allowed to be, and nothing more. Somebody with a 180 ms reaction and no idea where to look still loses.

What we do claim is narrower and, we think, true. These specific things are measurable, they respond to practice, they are quick to test, and having a real number for them is better than having a vague feeling. The rest is up to you.