DeviceBenchFree jitter, butterfly and drag click test
Three clicking techniques, each on its own record
Three techniques, three ten-second runs, three separate records: jitter, butterfly and drag clicking are different skills with different hardware costs, and one combined score would describe none of them. Drag mode adds the two readings the technique is actually about: how far the pointer traveled across the pad over the ten seconds, and the longest unbroken burst of contacts arriving less than 40 ms apart. That last figure comes with a limit worth stating up front: a browser is handed events with no source attached, so nothing here can tell a genuine drag click from a switch that has started chattering or from a script.
- 100% free
- No signup
- 3 modes, 3 records
- 10-second runs
- 40 ms burst detection
Lock the wrist, tense the forearm, and let the tremor that produces drive one finger on the button.
Nothing here can tell a drag click from a chattering switch or from a script, because all three arrive as pointer events a few milliseconds apart and the browser is given no source for any of them. If you did not touch the mouse and the counter still moved, the fault is the switch.
How to measure jitter, butterfly and drag clicking separately
Three runs rather than one, because the three techniques fail in three different ways.
Pick one technique and stay in it for the whole run
Each mode keeps its own record and each is a different skill, so mixing them inside one run produces a figure that belongs to neither. Jitter is forearm tension driving one finger. Butterfly is two fingers alternating on the same button. Drag is one stroke of a fingertip across the button. The card you pick tells you what it wants before the clock starts.
Run the full ten seconds, even when the technique collapses
All three modes use the same fixed window so the three records can be read against each other, and the second-by-second chart underneath is the reason not to stop early. Jitter almost always shows a cliff rather than a slope — the forearm holds the tremor for four or five seconds and then loses it — and that cliff is far more informative about whether the technique is usable in a real match than the average is.
In drag mode, read the burst and the travel, not the total
A drag run's total is the least interesting number on the screen. The longest burst counts the contacts that arrived under 40 ms apart in one unbroken sequence, which is the drag itself rather than the ordinary clicking around it, and the travel figure is how far the pointer moved across the pad over the ten seconds. Clicks per 100 pixels is the ratio that tells you whether the grip is working or whether you are just clicking quickly while moving your hand.
Technical specifications
| Modes | Three, each with its own stored record: jitter, butterfly and drag |
|---|---|
| Run length | Fixed at 10 seconds for all three, so the three records can be compared directly |
| Bounce threshold | 40 ms — the shared engine's cutoff for two presses too close together to be two deliberate ones |
| Longest burst | The longest unbroken sequence of gaps under that threshold, reported as a number of clicks |
| Drag travel | Pointer distance across the pad over the run, in CSS pixels, summed from uncoalesced pointer samples |
| Drag density | Clicks per 100 pixels of travel — the ratio that separates a working grip from a fast hand |
| Presses read from | pointerdown, not click, so nothing the browser synthesizes as a double click reaches the count |
| Cannot distinguish | A drag click, a chattering switch and a script — all three arrive as identical events with no source attached |
Frequently asked questions
What actually happens when someone drag clicks?
The switch is being made to bounce on purpose. Dragging a fingertip across the button applies pressure in a rapid stutter rather than one clean push, the contact inside the micro-switch makes and breaks several times during the stroke, and the mouse reports each of those as a separate press. It is the same physical event as the fault that makes an old mouse open a file twice — the difference is that here you are causing it deliberately, which is why the intervals land in a range no hand can produce.
Why does drag clicking work on some mice and not others?
Two things have to line up: grip and debounce. The button shell needs enough texture for a fingertip to stutter across rather than slide, which is why glossy top shells are almost impossible to drag on and why people tape their buttons. Then the mouse's own firmware has to pass the bounces through — many mice apply a debounce window of several milliseconds that deliberately swallows exactly this, and on those no technique will help, because the extra contacts are being discarded before they ever leave the device.
Will these techniques wear out my mouse?
Drag clicking will, and quickly by the standards of a mouse. A main-button micro-switch is typically rated somewhere between 10 and 20 million actuations, which under ordinary use is years of work; at the 25 or more presses a second a good drag produces, a 20-million switch is around nine days of continuous clicking. Butterfly clicking roughly doubles the rate of normal use, which is survivable. Jitter clicking barely touches the switch at all — its cost is paid by you rather than the hardware.
Does jitter clicking hurt?
It can, and the mechanism is not subtle: the technique works by holding the forearm under sustained tension so that the resulting tremor moves the finger, which is close to a definition of the loading that causes repetitive strain injury. Aching in the forearm or the base of the thumb during or after a run is the signal to stop, not to push through, and it is worth knowing that the technique is the one of the three with a real physical cost attached to practicing it.
Why is my butterfly score not double my one-finger score?
Because the limiting factor stops being how fast a finger can move and becomes how precisely two of them can take turns. Each finger only has to manage half the rate, which is why the technique works at all, but the two have to interleave cleanly — and when they drift toward landing together the button gets two presses inside its own bounce window and the mouse reports one. A butterfly run that reads barely above a normal run usually has a timing problem rather than a speed problem.
Will a server treat this as cheating?
Very possibly, and drag clicking specifically. Most large Minecraft servers publish a click-rate limit somewhere in the region of 15 to 20 clicks per second and explicitly treat drag clicking as autoclicking, because a server sees packets rather than fingers and a drag burst is indistinguishable from a macro. Jitter and butterfly clicking usually sit under the published limits, but the rate is what gets measured, not the intent, so a very good butterfly run can trip the same threshold.
The page says my gaps are under 40 milliseconds. Is my mouse faulty?
Not in drag mode — that is the technique doing its job, and a run with no gaps under 40 ms means the drag is not catching at all. The same reading in the middle of an ordinary run is a different story: nobody presses twice inside 40 ms deliberately, so if you were clicking normally and the counter flags bounces, the switch is chattering on its own and the mouse is on its way out.
About the three techniques and what each one costs
The three methods look similar on a leaderboard and have almost nothing in common underneath. Jitter clicking is a whole-arm technique: the forearm is held rigid, the tremor that tension produces is transmitted down to one finger, and the button is being pressed by a vibration rather than by a decision. Butterfly clicking is a coordination technique — two fingers alternating on the same button, each managing half the rate, so the ceiling is how cleanly they interleave rather than how fast either one moves. Drag clicking is not really a clicking technique at all: one stroke of a fingertip across the button surface stutters the contact inside the switch so that a single physical gesture is reported as a run of separate presses.
That difference decides who pays. Jitter clicking costs the hand and barely touches the mouse, and the cost is real enough to take seriously — sustained forearm tension to produce a tremor is close to a textbook description of the loading behind repetitive strain injury. Butterfly clicking roughly doubles the rate of ordinary use, which a switch rated for tens of millions of actuations absorbs without complaint. Drag clicking is the one that destroys hardware: at 25 or more contacts a second, a switch rated for 20 million presses has about nine days of continuous clicking in it, and it degrades long before that — the first symptom is the mouse starting to double-click on its own during normal use, which the mouse test will show on the button you have been dragging.
The third cost is the one people find out about after the ban. A game server never sees a finger; it sees clicks arriving in packets, and it makes its judgment from the rate and the spacing. Most large Minecraft servers publish a limit in the region of 15 to 20 clicks per second and treat drag clicking as an autoclicker outright, precisely because a drag burst has the interval signature of a macro rather than of a hand. That judgment is also made through your connection, so how those clicks are batched on the way out matters — a run measured cleanly here can look different at the far end if the link is unsteady, which the ping test will tell you before a match rather than after one. For a plain figure with no technique in it, tap normally on the CPS test, and for the same run on the second button there is the right click CPS test.
Where the three records live
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Nothing you type, paste or open is uploaded, logged or kept, which is also why the tools carry on working after you disconnect from the network.
Three numbers survive a reload — your best figure for each of the three techniques — and nothing else. They stay in this browser on this machine, no run is uploaded anywhere, and the control that clears all three sits next to the copy button.