| Jitter | Butterfly | Drag | |
|---|---|---|---|
| The motion | Tense the forearm until the hand shakes; the shake clicks | Two fingers on one button, alternating | Slide a finger across the button; friction makes the switch bounce |
| Typical speed | 10 to 14 CPS | 10 to 20 CPS | 20 to 40+ CPS in bursts |
| Lasts | Seconds, then the arm gives out | As long as your fingers do | One drag at a time |
| Needs | A light switch, a mouse you hold loosely | Any mouse | A grippy shell and a switch that bounces; most mice can't |
| Costs | Forearm strain, poor accuracy | Almost none, some learning | The switch, fast; anti-cheat flags |
| Timing signature | Fast, ragged gaps, no pattern | Short gap, long gap, short, long | Piles of gaps under 30 ms |
Jitter
Jitter is the oldest of the three and the least pleasant. You tense the muscles of the forearm against each other until the hand vibrates, and you rest the finger so lightly on the button that the vibration presses it. Good jitter clickers get 12 to 14 clicks a second for a few seconds. The clicks are fast but never regular, because a muscle tremor isn't a metronome. In the gap data it shows as a high rate with 10 to 40 percent spread and no alternating pattern, which is exactly what the jitter test looks for.
It hurts. Not immediately, but people who do a lot of it get forearm and wrist problems, and aiming while your arm is shaking is as hard as it sounds.
Butterfly
Two fingers, usually index and middle, on the same button, taking turns. Each finger is only clicking at 5 to 8 a second, which is easy, and together that is 10 to 16. It is the most sustainable of the three and the easiest to learn. Because the two fingers are never quite symmetric, the gaps alternate short and long, and the butterfly test measures how strongly. As you get better the alternation fades, which is fine: the detector then calls it "fast and even".
Some mice fire twice per press when butterfly clicked, because the second finger lands while the switch is still settling. Players look for this on purpose. Servers with anti-cheat look for it too.
Drag
Drag clicking isn't really clicking. You slide the fingertip across the button under light pressure, the friction makes the button vibrate, and the switch fires on every bounce. Twenty to forty clicks in one drag is normal on a mouse that supports it. Most mice don't: it needs a textured or slightly tacky shell and a switch that bounces rather than debounces, and manufacturers of ordinary mice work hard to prevent both. People add grip tape to get it.
In the data it is unmistakable: a third or more of the gaps are shorter than 30 ms, which no finger can do. That is the whole detector on the drag test. It wears the switch out in months, and almost every Minecraft server treats sustained rates this high as cheating.
Which one
If you want a higher number on a test, butterfly. If you want a higher number in a game, check the game's rules first; on most servers anything past 15 or so gets you kicked whichever way you produced it. And if the aim is to click a bit faster in everyday use, none of them: hovering and relaxing will get an ordinary finger to 8 or 9 with no side effects.